UNIVERSITY of CALIFORNIA Santa Barbara Wind

Home , Battle of Nagashino, Bunki, Bunmei, Echigo Province, Genki (era), Hi Province

UNIVERSITY OF CALIFORNIA

Santa Barbara

Wind, Forest, Fire, and Mountain 風林火山: The Evolution of Environmental

Management and Local Society in Central Japan, 1450-1650

A dissertation submitted in partial satisfaction of the

requirements for the degree Doctor of Philosophy

in History

John Elijah Bender

Committee in charge:

Professor Luke S. Roberts, Chair

Professor Anthony Barbieri-Low

Professor Peter S. Alagona

Professor Katherine Saltzman-Li

September 2017

The dissertation of John Elijah Bender is approved:

______Anthony Barbieri-Low

______Peter S. Alagona

______Katherine Saltzman-Li

______Luke S. Roberts, Committee Chair

August 2017

Wind, Forest, Fire, and Mountain 風林火山: The Evolution of Environmental

Management and Local Society in Central Japan, 1450-1650

John Elijah Bender

iii

ACKNOWLEDGEMENTS

Most scholarly acknowledgements pages, certainly the majority of those I have seen, thank family and dearest friends last. It has always seemed a bit odd to me, as this group comprises a core network of love, companionship, and support, without which none of my academic endeavors would have been possible. So I am going to break with tradition and begin by thanking my wife Lena. She is my best friend and fiercest advocate. I am every day grateful that she is there walking with me through this crazy journey called life. There are many other people I love and care for, who have contributed to the successful completion of this project. I will name but a few of the most influential, acknowledging that I am indebted to many more. My sister Annah, a brilliant and creative force, has always been in my life. She has been an unending source of advice, encouragement, friendship, and has profoundly affected by development as a person and as a scholar. She is one of the smartest people I know, probably the best writer to boot. My friend Ryan Abrecht has had a similarly broad impact. Ryan was one of first graduate school friends at the UCSB, and we later shared a home. He once drove me to the hospital in the middle of the night after a motorcycle accident, and he more recently performed my wedding. I am sure he got sick of hearing about water disputes and typhoons and so forth at some point, and will be very happy to know this thing is finally done. Some other cherished graduate school companions include Brian Tyrrell, a terrific person to hike around the Grand Canyon with, and good friend to stand in solidarity with while navigating the utterly terrifying academic job market; Hanni Halil, one of the most resilient and kindest people I know; Viktor Shamgin, my ever wise and generous senpai; and Travis Seifman, a thoughtful friend and talented fellow Japanese history nut. I also want to mention Nicole Pacino, Timothy Daniels, Munther Al-Sabbah, Tim Paulson, Jessica Marter-Kenyon, and Dusty Hoesely – all of whom were key players in the journey that has culminated in the completion of this project. I miss you all and very much look forward to future fellowship. My friend and mentor Luke Roberts has had a more direct impact than anyone else on the ideas laid out in the following pages. Thank you for years of patient, nurturing instruction. Luke taught me the basics from how to read a medieval text, to how to develop and express ideas as a scholar. My other advisers deserve a great deal of credit as well. Anthony Barbieri-Low was extremely helpful and supportive, Peter Alagona was responsible for introducing me to this whole thing called “environmental history,” and Katherine Saltzman-Li has been a stalwart source of encouragement, keen advice, and incisive feedback. I have learned a great deal from each of you, and am forever indebted for all you invested in me as teachers. And, thanks in no small part to you, I will not be requesting another round of letters this fall. I want to extend a very special thanks to Professor Sasamoto Shōji, who acted as a sponsor during a research year in Japan on a Fulbright scholarship. I was, and continued to be, completely awed by Professor Sasamoto’s breadth of knowledge. But even more impressive is how he is able to weave so many complex, seemingly disparate elements into fascinating analysis. The way Professor Sasamoto poses historical questions left a big impression, and I will continue to ponder the many lessons

imparted for years to come. Thank you for your incredible generosity, which was almost singularly responsible for such an enjoyable, productive year in Matsumoto. Regarding that year abroad, I owe a huge debt to the people at the Japan-US Educational Commission, who administer Fulbright funds. Jinko Brinkman in particular was incredibly helpful, and without her logistical support, I would not have been able to enjoy a year of study at Shinshū. Thank you as well to Kate McDonald, Sabine Fruhstuck, and Paul Spickard – all terrific teachers who took the time to offer guidance even though there was no obligation to do so. And, going back to the days when my knowledge of Japan and understanding of history was yet in infancy, thank you to Wayne Farris and Dominic Capeci. I have had the great fortune of sharing ideas and learning from this talented group of experts. There are so many more influential teachers, simply too many to name. Finally, I wish to dedicate this work to my daughter Danika. Daddy finished what for him has been this monumental undertaking just a few weeks before you were born, but it has not seemed quite as important since you have been here. I cannot wait to talk ideas with you someday soon.

Thank you all,

Elijah Bender August 2017

VITA OF JOHN ELIJAH BENDER August 2017

EDUCATION

Bachelor of Arts in History, Missouri State University, May 2005 (cum laude) Master of Arts in History, University of Hawaii at Manoa, May 2008 Doctor of Philosophy in History, University of California, Santa Barbara, September 2017 (expected)

PROFESSIONAL EMPLOYMENT

2005-2008: Teaching Assistant, Department of History, University of Hawaii at Manoa 2011-2017: Teaching Assistant, Department of History, University of California, Santa Barbara 2017-present: Assistant Professor, History Department, Concordia College

MASTERS THESIS

“The Last Man Standing: Causes of Daimyo Survival in Sixteenth Century Japan”

AWARDS

Foreign Language and Area Studies Fellowship, summer 2006 University of California, Santa Barbara Special Regent’s Fellowship, 2010-2015 Japan Foundation Graduate Student Fellowship, spring 2012, spring 2013 Richard and Jeanne Williams Endowed Graduate Fellowship, summer 2013 University of California, Santa Barbara History Associates Graduate Fellowship, summer 2013 Japan Foundation Dissertation Research Abroad Fellowship, 2014-2015 (declined in favor of Fulbright) Fulbright Institute of International Education Graduate Research Fellowship (Japan), 2014-2015

FIELDS OF STUDY

Major Field: premodern Japanese history

Studies in Japanese history with Professors Luke Roberts, Kate McDonald, and Katherine Saltzman-Li

Studies in Chinese history with Professor Anthony Barbieri-Low

Studies in World history with Professors Harold Marcuse and Paul Spickard

Studies in Environmental history with Professor Peter S. Alagona

ABSTRACT

Wind, Forest, Fire, and Mountain 風林火山: The Evolution of Environmental

Management and Local Society in Central Japan, 1450-1650

John Elijah Bender

What role did contests over environment play in the large scale transition from war to peace during Japan’s long sixteenth century? This study examines the question by focusing upon how people dealt with environmental challenges and disputes during an age of instability, institutional failure, and war. Japan’s late medieval (ca. 1450-1600) environmental management regime underwent important changes that proved crucial in ending a long period of warfare and formed the basis of an even longer era of stability. The construction of stable patterns of environmental management depended upon a significant increase in de facto local autonomy, coupled with the rise of authoritarian regional states. Although these developments sometimes stood in opposition to one another, they ultimately combined to reorganize the environmental management regime in Kai and Shinano Provinces.

Central Japan became a crossroads of multiple struggles during the late medieval period. The Takeda warrior house of Kai Province eventually emerged from a pool of contenders to claim supremacy in the region in the first few decades of the sixteenth

vii

century. Like other regional magnates (daimyo), the Takeda had an interest in ending warfare and tapping into local productivity within their territory. They did so in ways that effectively turned the Takeda into local patrons who had the means to act as a guarantor of local interests. Communities had been largely forced to fend for themselves during the civil war, or Warring States, era (1467-1600). As a result, they developed administrative and defensive procedures independently. The Takeda worked to integrate these communities by leaving most of those local practices in place, deputizing a class of officials, and negotiating set tax rates on local production.

Residents did surrender some of their prerogatives, but collectively gained a more favorable institutional framework. Takeda and their officials were able to mediate local disputes, something that previously could only be done by two parties directly, often resulting in violence. Crucially, the Takeda possessed the wherewithal to enforce dispute settlements. They also supported regional infrastructure projects and policies that required various communities to coordinate certain management practices. These projects ultimately increased efficiency, and linked villages in ways that created a set of complimentary interests between different socioeconomic classes.

The development of complimentary interests in turn created a vested interest in maintaining new sociopolitical relationships. As these bonds strengthened, they became more regularized, more reliable, and more stable. This gave added momentum towards the reestablishment of order at the close of the sixteenth century. The local settlements worked out at this time remained the basis of local administration well into the early modern period. Forged as they were over generations of upheaval, these practices favored maintenance of stability over other concerns. Often, that entailed relatively

viii

hands-off treatment from political superiors. The flexibility between the authoritarian discourse of samurai rule, and the actual latitude afforded local communities in their affairs, significantly contributed to the remarkable durability of the Tokugawa system.

TABLE OF CONTENTS

Introduction: Environment and Premodern Political Economy in the Japanese Islands...1 A. Defining Civil War History – History Defining Place……………………………………13 B. Evolution from the Tributary to Market to Commercial Economy……………..…24 I. From a Warring State to Warring States: The Collapse of the Estate System and the End of Stability in Fifteenth Century Kai...... 33 A. Paradigms of Dispute in Medieval Japan………………………………………………….....39 B. Tensions with the Center-Provincial Political Economy………………………………46 C. Uesugi Zenshū’s Rebellion and Instability in Kai…………………………………………54 D. Conclusion: Administrative Anachronism…………………………………………………..63 II. Fighting Over Nature: Resource Disputes in an Age of Instability…………………………...70 A. People and Place of Late Medieval Yoshida…………………………..…………………….81 B. Case 1: Reactions to Local Infrastructure Building (Yoshida vs. Kawaguchi, 1475; Yoshida vs. Kobayashi, 1504)……………………………………………...…………...91 C. Case 2: Intervention in Infancy (Yoshida vs. Watanabe, 1533)……….……………98 D. Case 3: Disorder and Procedure (Yoshida vs. Kobayashi, 1556-57)………..….102 E. Case 4: Local “Ownership” Confirmed (Yoshida vs. Kobayashi, 1559)…...……114 F. Case 5: Judicial Resolution (Yoshida vs. Kurechi and Onuma, 1635)………..…118 G. Conclusions: Civil War-era Disputes in Context…………………..…………………….124 III. Subjugation to Stability: Building Complimentary Interests across Socioeconomic Levels………………………………………………………………………………………………………………131 A. From Province to Country: Incorporation………..……………………………………….133 B. Reorganization of Service Obligations…………………………..………………………….147 C. Construction Duty in late Medieval Kai……………………..……………………………...159 D. Conclusions: Gaining a Stake in the New Order……………………………..………….174 IV. Storming into Exile: Disaster, Embedded Tension, and the Complex Historical Legacy of Calamity……………..……………………………………………………..…………………………………179 A. A Hundred Year Storm…………………………………………………….…………………...... 185 B. Medieval Disaster Response and Debt Relief in Kai…………….……………………189 C. Embedded Tensions in Late Medieval Society…………………….……………………200 D. Implications – Do Disasters Matter?...... 215 Conclusion: Perspectives on Environment and Order……………………………………………...222 Bibliography………………………………………………………………………………………………………….228 Appendix………………………………………………………………………………………………………………235

LIST OF MAPS AND FIGURES

Maps Map x.1: Premodern Japanese Provinces…………………………………………………………………….7 Map x.2: Major Daimyo of the Late Sixteenth Century………………………………………………..19 Map 2.1: Regions of Kai Province (Yamanashi Prefecture)…………………………………………78 Map 2.2: Yoshida and Environs………………………………………………………………………………...85 Map 2.3: Key Sites in Late Medieval Yoshida……………………………………………………………..90 Map 2.4: Key Sites of the 1635 Dispute…………………………………………………………………...119 Map 3.1: Gold Mines of Kai Province……………………………………………………………………….148 Map 3.2: Shingen’s Dike and Environs…………………………………………………………………….164 Map 3.3: Villages Involved in the Ryūō River Works………………………………………………..166 Map 4.1: Lake Kawaguchi and Environs………………………………………………………………….186 Map 4.2: Kai-Shinano Border Area………………………………………………………………………….188

Figures Figure x.a: Takeda Shingen Statue in Kōfu………………………………………………………………...15 Figure x.b: Takeda Shrine………………………………………………………………………………………...17 Figure x.c: Shingen’s War Banner……………………………………………………………………………..18 Figure 1.a: Status Groups and Office of Medieval kenmon…………………………………………..38 Figure 1.b: Economic and Political Relationships of the Estate System……………………..235 Figure 1.c: Economic and Political Relationships of Provincial Land………………………...236

Introduction: Environment and Premodern Political Economy in the Japanese Islands

Contemporary Japan is one of the safest places on earth. Visitors to the island nation often remark upon its cleanliness, order, and extremely low crime rates. Yet over the long course of Japanese history, contemporary society is an acute outlier. With the exception of the remarkably stable early modern and post-1945 eras, most of Japan’s history has been punctuated by persistent warfare and social strife. No era embodies this trend more clearly than the Warring States Period (ca. 1450-1600).1 Over one hundred and fifty years of civil war raged. Violence was endemic, afflicting all aspects of daily life for those who lived through the conflict. And then, in the late sixteenth century, several decades of rapid, sweeping sociopolitical changes ended generations of bloodshed. When swords were sheathed and arrows lowered a new order emerged.2

Consistent upheaval gave way to stability, violent conflict replaced by bureaucracy. This new order proved remarkably durable, lasting as it did without major incident for over two and a half centuries.

1 Dating the Warring States period is subject to debate. Traditionally, it is considered to have begun with the outbreak of the Ōnin War in 1467 and ended when the warlord Oda Nobugawa took Kyoto in 1568. However, widespread warfare continued throughout the so-called “unification” era (1568-1590), and was not wholly finished until the decisive Battle of Sekigahara in 1600. Another potential conflict was not fully resolved until the Osaka campaign of 1614-15. In Kai and Shinano, civil war raged from Uesugi Zenshū’s Rebellion of 1417 continuously until the Tokugawa invasion of 1582. This study uses the terms “Warring States,” “Sengoku,” “Civil War Era,” and “late medieval” interchangeably to denote the years of upheaval from roughly 1450-1600. My use of these dates and the labels employed are discussed in subsequent chapters. 2 The “Great Peace” of the Tokugawa Period (1600-1868) endured for some two hundred and sixty eight years. I refer to the “Tokugawa,” “early modern” or “Edo Period” order to describe the sociopolitical system in place during these centuries which maintained the “Great Peace.”

The transition from longstanding war of the late medieval era to the extended peace of early modern Japan has fascinated generations of scholars. Just how did this transition occur? Researchers who examine the transition from an archipelago wide perspective have tended to emphasize elite political and military drama. For them, it was a new class of regional magnates (warlords or daimyo 大名) who forged the early modern order via conquest and decree.3 In the estimation of some, the most significant warlord achievement was the subjugation of the professional warrior class. Restriction of autonomy and bringing unruly samurai to heel opened the door for peace and the new social order.4 Another strand of scholarship has emphasized elite and local interaction as key. They perceive more input and influence from the bottom of the socioeconomic hierarchy than scholars in the previous group. Struggles between rulers and ruled over taxes, labor, and governing authority comprise the most important issues.5 These tumultuous negotiations eventually led to new hierarchical arrangements

3 This was typical of institutional historians who flourished in the 1970s and 80s, when studies of the late medieval-early modern transition became more numerous. Nonetheless, notable contemporary scholars continue to emphasize conquest and control of territory. Representative works of major scholars include: John Whitney Hall, Government and Local Power in Japan, 500 to 1700 (Princeton: Princeton University Press, 1966); John W. Hall, Nagahara Keiji, and Yamamura Kōzō, eds., Japan Before Tokugawa: Political Consolidation and Economic Growth, 1500-1650 (New Jersey: Princeton University Press, 1981); Takagi Shōsaku, “Hideyoshi’s Peace and the Transformation of the Bushi Class: The Dissolution of the Autonomy of the Medieval Bushi,” Acta Asiatica: Bulletin of the Institute of Eastern Culture 49 (1985): 46–77; Fujiki Hisashi, Toyotomi heiwarei to sengoku shakai (Tokyo: Tokyo daigaku shuppankai, 1985); Shibatsuji Shunroku, Sengoku daimyō-ryō no kenkyū: Kai Takeda-shi ryō no tenkai (Tokyo: Meicho Shuppan, 1981). 4 Scholarship of this type flourished in the 1980s and 90s in Japan, and was reflected in Western languages around the same time. In English, several contemporary scholars focus upon the late medieval transformation of the samurai class. See for example Michael P. Birt, “Samurai in Passage: The Transformation of the Sixteenth-Century Kanto,” Journal of Japanese Studies 11, no. 2 (July 1985): 369–99; Eiko Ikegami, The Taming of the Samurai: Honorific Individualism and the Making of Modern Japan (Cambridge: Harvard University Press, 1995); David Anthony Eason, “The Culture of Disputes in Early Modern Japan, 1550-1700” (PhD diss., UCLA, 2009); David Spafford, A Sense of Place: The Political Landscape in Late Medieval Japan (Cambridge: Harvard University Press, 2013). 5 Emphasis upon a “give and take” between elites and their subjects advanced in the 1990s and continues to be a prominent trend in contemporary scholarship. Some key examples include Fujiki Hisashi, Sengoku no sahō: mura no funsō kaiketsu (Tokyo: Heibonsha, 1987); Kōzō Yamamura, “From Coins to Rice: Hypotheses on the Kandaka and Kokudaka Systems,” Journal of Japanese

which proved resilient enough to bring peace. More recently, there has been a trend towards analyzing the widespread changes of the late medieval era from the perspective of local society. These scholars have pinpointed development as the key agent of change.6 While both the “negotiation” and “development” bodies of research centers the local, non-elite, the work primarily analyzes changes in local society as resulting from regional warlords’ state-building policies. Although all of the scholarship cited here represents a huge variety of opinion and diverse analysis, it generally prioritizes the institutional, political, and technological.

Social histories of late medieval Japan are fewer in number, due largely to the availability of desired evidence, but have proven influential. The first notable social historian of the period was Amino Yoshihiko.7 Amino drew upon non-traditional sources such as folklore to attempt to reconstruct a picture of social life in premodern times. He called for an emphasis upon local practice, arguing that it was changes here

Studies 14, no. 2 (July 1988): 341–67; Keiji Nagahara and Kōzō Yamamura, “Shaping the Process of Unification: Technological Progress in Sixteenth- and Seventeenth-Century Japan,” Journal of Japanese Studies 14, no. 1 (January 1988): 77–109; Birt, “Samurai in Passage”; Philip C. Brown, Central Authority and Local Autonomy in the Formation of Early Modern Japan: The Case of Kaga Domain (Stanford: Stanford University Press, 1993). Note that Fujiki and Birt take a kind of middle stance between top-down and bottom-up socio-institutional history. Both make compelling arguments for the role of non-elites in the transformation of society. However, they view this role as a reaction to new impositions by territorial rulers. 6 Development here includes technological change, expansion of arable acreage, more intensive or efficient resource extraction, infrastructure building, urban growth, and building construction. Some representative works dealing with the region and time period under study here are Osano Asako, “Takeda ryōgoku-ka ni okeru kokujin Oyamada-shi to gunnai no zaichi seiryoku,” Takeda-shi kenkyū 30 (2004); Osano Asako, “Takeda ryōgoku no dogōsō to chiiki shakai,” in Sengoku daimyō Takeda- shi no kenryoku to shihai, ed. Hirayama Masaru and Marushima Kazuhiro (Tokyo: Iwada shoin, 2008); Nishikawa Kōhei, Chūsei kōki no kaihatsu – kankyō to chiiki shakai (Tokyo: Kōshi shoin, 2012). 7 Amino’s most famous work is undoubtedly Amino Yoshihiko, Muen, kugai, raku: Nihon chūsei no jiyū to heiwa (Tokyo: Heibonsha, 1987). Amino published prolifically during a long career and influenced a generation of scholars. For a representative summary of his major historical arguments, see Amino Yoshihiko, Rethinking Japanese History, trans. Alan S. Christy (Ann Arbor: University of Michigan Press, 2012).

which ultimately transformed society. Many scholars have greatly expanded upon

Amino’s ideas. Their chief insight into late medieval Japan is that warlords’ seemingly new policies were often reactions to existing local political and economic arrangements.8

The varied perspectives described above have resulted in many compelling arguments, yet the role of environment in the widespread sociopolitical changes of late medieval Japan has not been adequately examined. Focusing on environmental management allows us to avoid an oversimplified binary dividing premodern society into classes of ruler and ruled. Elite conquest, struggles over taxes and the terms of submission, local disputes over construction or land use, and fights to mitigate famine or disease were surely influential upon the transition from war to peace. However, these contests were symptoms, manifestations, of a more fundamental conflict concerning environmental management. Japan’s late medieval evolution was not the result of policies imposed upon the region by elites. Rather, solving local environmental management problems was key. It was these issues, these arenas of conflict, which drove the evolution of local society. Communities found themselves at odds with other local actors as often as with remote authorities. This process therefore consisted not merely of negotiating with, resisting, or submitting to elites. Those examples do exist,

8 Social histories of late medieval Japan have increased in number since the 2000s. In that field, the most influential scholars to this project are Sasamoto Shōji and Wayne Farris. Sasamoto’s work is characterized by an emphasis upon local initiative shaping daimyo policy, rather than the other way around. Farris examines the late medieval transition primarily from the perspective of demographic growth, which he argues was driven by technological change and increased resistance to epidemic disease. See Sasamoto Shōji, Sengoku daimyō Takeda-shi no kenkyū (Kyoto: Shibunkaku Shuppan, 1993); William Wayne Farris, Japan’s Medieval Population: Famine, Fertility, and Warfare in a Transformative Age (Honolulu: University of Hawai’i Press, 2006).

most notably in the form of regional leagues (ikki 一揆). Nonetheless, in many cases, local residents employed elites to advance their own agendas vis-à-vis nearby competitors or even between factions of their own community. So much of Warring

States conflict was by necessity between locals. Since these were chiefly environmental problems, analyzing local environmental management can help expose the root causes of disorder. As we will see, rulers often had to adjust to local circumstances, maintaining established practices and endorsing local management preferences.

Premodern political economy was therefore not solely or even principally a struggle between rulers and ruled. Take, for example, the quintessential elite establishment of a new sociopolitical order and one that marks the boundary between medieval and early modern: conqueror and “unifier” Toyotomi Hideyoshi’s 豊臣秀吉 (1537-1598) “realm at peace” edicts (sōbujirei 総無事令).9 These orders were not imposed upon a society that was then transformed. As this study hopes to demonstrate, Hideyoshi’s edicts and similar ones of regional rulers were actually the culmination of longstanding settlements that had been worked out at the local level over several decades during the civil war era.

In the late medieval era, significant changes in local environmental management practices proved crucial in ending a long period of warfare and formed the basis of an even longer era of stability. The construction of stable patterns of environmental management depended upon a significant increase in de facto local autonomy, coupled with the rise of authoritarian regional states. Although these developments sometimes

9 For a summary of the current scholarly debates over sōbujirei, see Marushima Kazuhiro, Sengoku daimyō no gaikō (Tokyo: Kōdansha, 2013).

stood in opposition to one another, they ultimately combined to reorganize society. In the civil war context, local people were forced to manage affairs and solve their own problems. Regional elites gradually realized that in order to incorporate territory it was expedient to respect local decisions and largely leave local practices in place. This created an incentive to recognize warlord government as the legitimate peace-keeping authority. Slowly, progress towards stability gained momentum, accelerating in the final decades of the sixteenth century. In order to understand how this process actually took place, we must closely examine local environmental management in specific locations and trace its evolution. This study takes as its object the province of Kai 甲斐

国 in eastern central Japan (see map x.1).10 The historical circumstances of Kai make it a useful tool for exploring central-regional and regional-local tensions in the medieval era

(ca. 1150-1600), which were at their core contests over environmental management.

Central Japan became a crossroads of multiple struggles during the late medieval period. The Takeda 武田 warrior house of Kai eventually emerged from a pool of contenders to claim supremacy in the region in the first few decades of the sixteenth century. Like other regional magnates, the Takeda had an interest in ending warfare and tapping into local productivity within their territory. They did so in ways that effectively turned the Takeda into local patrons who had the means to act as a guarantor of local interests. Communities had been forced to fend for themselves during the civil war. As a result, these settlements had developed their own economic, administrative, and military practices. They came to control, or fought for control of,

10 Modern Yamanashi Prefecture 山梨県.

Map x.1: Premodern Japanese Provinces

their local environment independently. The Takeda worked to integrate such communities by leaving most of those local practices in place, deputizing a class of officials, and negotiating set tax rates on local production. Residents did surrender some of their prerogatives, but collectively gained a more favorable institutional framework. Takeda and their officials were able to mediate local disputes, something that previously could only be done by two parties directly, often resulting in violence.

Crucially, the Takeda possessed the wherewithal to enforce dispute settlements. They also supported regional infrastructure projects and policies that required various communities to coordinate certain management practices. These projects ultimately increased efficiency, and linked villages in ways that created a shared set of interests.

This study relies heavily upon several terms and concepts which are inherently anachronistic. The most basic is the word environment itself. This was not a term employed contemporaneously, nor does it correspond neatly to the ways that late medieval people understood the places around them.11 Premodern peoples thought about their environment in terms of specific activities and resources. There were places for farming, grazing, gathering lumber, streams for fishing, waterways for transport, and “undeveloped” land attached to one settlement or another. In this way the environment, in the contemporary sense, was compartmentalized into specific components. It follows that there would not have been a concept of overall management of these things.

11 The contemporary Japanese word for environment is kankyō 環境, but was not used in premodern times. The term shizen 自然 (“nature”) was used to broadly refer to phenomena external to human society, including the physical world, biota, animals, and the many spirits that populated these “natural” places. Humans and shizen were therefore separate entities which interacted in a number of ways, but did not mutually shape one another.

For our purposes here, environment is defined as a network of socio-ecological relationships between the human community and non-human actors in a particular place. Environmental management is employed to denote the social, economic, and political practices which dictate the human activities of those relationships. As a consequence, environmental management affects all aspects of society to some degree.

It is why contests over local resources were so central to Warring State period conflict.

Without stable access to necessary resources, recognition of claims to a particular place, acceptance of borders and definitions of space, or mechanisms for resolving disputes over these issues, there could be no lasting peace. Because conflicts over environmental management present such fundamental problems for society at large, it took many years to find workable solutions in late medieval Japan. The tumult of the era added unpredictability and undoubtedly prolonged the process. But once a solution was in place, it created an incentive to keep it in place rather than risk descending back into the chaos of war. This is described as common interest here – motivation to maintain a set of relationships which were preferable to (possible) alternatives, such as organized violence. As more and more local areas arrived at such solutions, a greater portion of the sociopolitical hierarchy gained a vested interest in the emerging early modern order. We can see this process unfold in late medieval Kai.

This study opens with a macroeconomic context that helps explain regional conditions in Kai. Across the archipelago, numerous factors contributed to a greater concentration of resources and wealth in the provinces. The growth of regional markets and increasing prerogatives claimed by local landlords indicate this shift. As local economic output grew, competition for control of it intensified. This upset the

prescribed center-periphery balance of the ruling polity and led to serious political dysfunction. Repeated upheavals then catalyzed widespread collapse and war. The countryside was atomized between a number of different players including small-time warrior-landlords, religious institutions, and independent settlements.

In a highly competitive, fragmented environment, the resolution of resource disputes proved unpredictable. By examining accounts left behind by local monks, my second chapter shows that residents of rural settlements in Kai took active steps to protect their interests. Competing claims were dealt with via heavy-handed negotiations backed by militias comprised of the fighting-aged young men of the community. Villagers fought samurai armies and other militias alike. Pacification of territory by regional warlords thus depended upon more than defeat of rival armies.

Daimyo had to act as guarantors in order to ensure cooperation. Takeda investiture documents indicate the give and take arrangements employed to integrate communities into the domain. Locals received guarantees of specific rights and privileges in exchange for a percentage of their production. Within the community, this relationship reorganized hierarchies and led to a less participatory but largely autonomous administrative system.

Chapter three examines the creation of a new vested interest in the emerging late sixteenth century sociopolitical order for people of very different status groups.

Common interest in the maintenance of this order across the sociopolitical hierarchy was possibly the most influential impact of stable resource access in late sixteenth century Kai. It allowed for previously unachievable cooperation, such as that indicated

by large infrastructure projects. One famous example is Shingen’s dike in modern Kōfu

City 甲府市. The project represents the culmination of several trends explored in previous chapters. Rural settlements could rely upon the growing institutional structure of the domain to protect their claims. Furthermore, they saw that their material and physical contributions would be standardized. This meant that communities could rely upon the daimyo to organize inter-local projects, ensuring that each group contributed its agreed upon share. Increased cooperation in turn reinforced a common interest, providing further momentum towards the emerging early modern order. The benefits of stability became ever more apparent as time went on, and served as the guiding principle for elite administration. Even after the removal of the ruling

Takeda house from Kai, their successors (none other than the Tokugawa 徳川, who completed the project of bringing all Warring States daimyo under the influence of a single leader and became the ruling dynasty of the early modern age) recognized the utility of leaving local administrative settlements in place. Many such practices persisted well into the early modern period with little or no alteration.

Finally, in chapter four I consider unpredictable forces which could alter historical trends in complex ways. Although responsible for short term disruptions, natural disaster also facilitated momentum towards cohesion. Disaster was not an agent of structural change and ultimately reinforced the Takeda-led domain-building project in

Kai. The events of a 1540 coup, which unfolded as the result of a large typhoon, illustrate a tipping point between a longstanding trend of disorder towards order.

Increased stability in the countryside gave regional magnates greater influence in the

second half of the sixteenth century. This in turn reinforced their ability to mediate local disputes, thus advancing incrementally towards the reestablishment of order. But that process was continuously threatened by external rivals and internal disunity.

Underlying tensions between samurai, daimyo, rural residents, and townsfolk erupted in moments of crisis. In the civil war context natural disaster had a tendency provoke crisis with greater frequency than it would have in a comparatively calm age. The disaster of 1540 did just that, leading to the downfall of a very successful regional warlord. Yet although it stressed a fragile political order, these events also contributed to an uncommon cohesion amongst Takeda retainers, collectively the regional political elite. Takeda retainers were able to depose of a lord who fought too many battles during times of environmental stress, and replace with a leader more attentive to local interest without significant conflict. As a result, violence was avoided amidst a characteristically violent age.

This regional study contributes to a more nuanced understanding of a remarkable era in Japanese history. The transition from over a century of civil war to a period of unprecedented peace is a dramatic example of a shift from disorder to order. My goal is twofold: to provide a fresh analysis and new information on the environmental history of late medieval Kai Province, and to elucidate the influential role that contests over environmental management played in Japan’s late medieval to early modern transition.

It is my hope that this work will ultimately contribute to our understanding of how stability is formed and maintained in society, as well as the principle causes of disorder and violence. I begin with an introduction to Kai and central-regional tension so influential in the premodern era.

A. Defining Civil War History – History Defining Place

Endemic violence coupled with the total failure of central government institutions defines the civil war era. While this was an archipelago-wide phenomenon, several places are much more closely associated with this history than others. That is largely because of the traditional civil war history narrative. That story focuses attention upon the rise and fall of numerous territorial magnates. Epic clashes between samurai armies of one lord against another are key events, and the story ends when the strongest of those lords finally vanquished their foes and ruled Japan unchallenged. In this version of civil war history, the main players are warriors, and the main type of conflict is a military one over the acquisition of territory. At stake was territory and the means to rule, a kind of imperial drama. Kai Province plays a central role in that story. The problem, for historians at least, is that key elements of this imperial drama have been shown to be inaccurate.

The traditional story of the Warring States is important in present day Yamanashi

Prefecture. Sengoku Era history is so celebrated in part because the place is rather unremarkable contemporary Japan. Disparaging commentators might remark that it is

“nothing but mountains,” a play on the prefecture’s name.12 Other than mountains, it is known mainly for grapes, wine, udon, and one other thing that overshadows all the rest.

Yamanashi is remembered today as the land of the Takeda. Though this warrior lineage spanned many generations and multiple historical eras, the Takeda are most famous as

12 In Japanese, yama shika nashi 山しか無し means “nothing but mountains” and is sometimes applied by both locals and outsiders as an acknowledgement of Yamanashi’s rather pedestrian, peripheral role in contemporary Japanese culture and economy.

regional warlords during the civil war era.13 They have become one of the most celebrated Sengoku daimyo in all of Japan, and contemporary Yamanashi abounds with physical reminders of Takeda glory days in everything from monuments to t-shirts.

Just outside the train station in the prefectural capital Kōfu 甲府 stands an imposing stone likeness of the greatest of all Takeda lords, Shingen (1521-1573) 信玄. 14 In full armor, the warlord sits upon a commander’s field stool, his battle fan and longsword held prominently before him (see Figure x.a). Only a short walk from this twenty foot tall figure lies Takeda shrine 武田神社. A popular attraction, the shrine was once the site of the house seat, known as Tsutsujigasaki 躑躅ヶ埼 (Figure x.b). Shingen’s famous war banner flies prominently about the place, emblazoned with the famous

“fūrinkazan” 風林火山 slogan (figure x.c). Drawn from an ancient military text attributed to Sun Tzu 孫子 (544-496 BCE), the banner reads “swift as the wind, silent as the forest, fierce as fire, immovable as a mountain.”15 This phrase has entered the

Japanese lexicon and is associated with strength, wisdom, decisiveness, and cunning – an ideal combination of virtues in a leader. It evokes the rugged, powerful attributes of

13 In order to avoid repetition, I use the terms civil war era, late medieval era, Sengoku period (sengoku jidai 戦国時代), and Warring States period interchangeably to refer to the centuries of disorder from roughly 1450 to 1600. 14 Takeda Shingen presided over the greatest expanse of Takeda territory during the Warring States era. His domain peaked in the early 1570s and encompassed all of Kai Province, much of Shinano, and parts of Suruga 駿河, Kōzuke 上野, Tōtomi 遠江, Mikawa 三河, and Mino 美濃 Provinces (see maps x.1 and x.2). Known through much of his adult life as Harunobu 晴信, he took the Buddhist name Shingen in 1559. I have chosen to consistently refer to him as Shingen throughout the text to avoid confusion. For analysis of Shingen’s name change, see Hirayama Masaru, Kawanakajima no tatakai (senshi dokyumento), jō (Tokyo: Gakushū kenkyūsha, 2002), 154. 15 The phrase originates in the sixth century BCE treatise The Art of War 孫子兵法 (Sonshi heihō).

Figure x.a: Takeda Shingen Statue in Kōfu16

the landscape and inscribes them upon the region’s most famous hero. Shingen’s fame has forged a close association between place and the story of the Takeda in the late medieval era. Yamanashi’s identity is therefore deeply intertwined with the civil war era, Takeda warlords, and the striking environment that was their home.

16 Google images.

It was the warlord Shingen who led the Takeda army in some of the most famous battles of the era. The five clashes with the Uesugi 上杉 of Echigo 越後 in 1554, 55, 57,

61, and 64 at Kawanakajima 川中島 have become emblematic of the era (see map x.2).17

It was Shingen who bested the Tokugawa at Mikatagahara 三方ヶ原 in 1573, and were it not for his death only a few months later, scholars speculate the Takeda could have challenged the Oda 織田 for national supremacy with a move on Kyoto. Instead, Takeda fortunes soured with a devastating defeat at Nagashino 長篠 in 1575. Shingen’s heir

Katsuyori 勝頼 (1546-1582) never recovered his family’s former glory. Surrounded by enemies, he and a handful of retainers loyal to the bitter end committed suicide at

Tenmokuzan 天目山 in 1582.18

In the traditional telling of Takeda downfall, the clash with Oda Nobunaga 織田信長

(1534-1582), the first to make significant progress towards bringing rival daimyo into a coalition under his suzerainty, in 1575 represents a turning point. It is infused with the pathos of Ivan Morris’ tragic hero, played by Katsuyori, as well as a lament for the disappearance of traditional ways of waging war.19 The noble samurai, here imagined as the famed Takeda cavalry, succumbed to awesome yet brutish power of unskilled foot soldiers wielding muskets. The story goes that Katsuyori, thrust into a position of leadership after his father’s untimely death, marched out to meet this dangerous enemy

17 The Kawanakajima battle sites now lie within Nagano City 長野市, Nagano. The site of the most famous and consequential fourth battle in 1561 is now a park and home to the Nagano City Museum 長野市立博物館. 18 Katsuyori and retainers are enshrined at Keitoku-in Temple 景徳院 in Kōshū City 甲州市, Yamanashi. 19 Ivan Morris, The Nobility of Failure: Tragic Heroes in the History of Japan (Tokyo: Kurodahan Press, 1975)

Figure x.b: Takeda Shrine20

near a small castle in Mino Province in early summer (see Map x.1). Much rode on the battle, and the participants knew it. Victory would determine who held the upper hand in protracted back and forth between the two sides. The combined Oda-Tokugawa force threw all their might at a Takeda foe who had stymied them for nearly a decade. Using the innovative “three shot volley,” Oda musketeers mowed down a stupendous charge

20 Photo by author.

from the Takeda horse and won the day.21 Men and beast fell by the hundreds before the deadly new weapons. The engagement came to symbolize the coming of a new era. The old ways of war gave way to the new. Trained, professional, horse-riding samurai soundly defeated by ranks of foot soldiers wielding guns. A famous early seventeenth century screen painting depicted the battle in precisely these terms.

Figure x.c: Shingen’s War Banner The above account of Nagashino is one

of the most well-known episodes of entire

late premodern age. It also happens to be

one of many civil war battle accounts of

dubious historical veracity. Katsuyori’s

supposedly epic defeat, moreover, was

probably not so historically influential,

either.22 So much of Yamanashi’s Sengoku

history is part of a narrative that never

actually existed. Battles between samurai armies were part and parcel of late medieval history. It is no surprise that they have received so much attention, from professionals and enthusiasts alike. But their importance as historical agents has and continues to be vastly overestimated. These

21 The three shot volley was a gunnery tactic supposedly developed and implemented by Oda Nobunaga. A front line of musketeers took shots from their knees at an approaching enemy while two lines stood behind them to reload. With each volley, the front line stood up and retreated to the back while the next line replaced them. The historical veracity of this tactic is dubious, likely a dramatic invention of later generations of storytellers. See Jeroen Pieter Lamers, Japonius Tyrannus: The Japanese Warlord, Oda Nobunaga Reconsidered (Leiden: Hotei Publishing, 2000), 41. 22 Sasamoto, Sengoku daimyō.

Map x.2: Major Daimyo of the Late Sixteenth Century23

clashes were not even the most prevalent types of armed conflict during the “age of the country at war.”

The subject of disputes and violence in late medieval Japan has received ample attention from scholars. Studies have tended to focus upon hereditary, professional warriors and their struggles for territorial supremacy at various scales. Battles and skirmishes between fighting men are the most clearly visible examples of civil war era conflict. Territorial conquest, in particular at the regional level, has long been a focus of late medieval historical studies. No doubt the acquisition of territory by one warlord or another had its impact, yet it did little to structurally change society. What did change

23 Hall, Nagahara Keiji, and Yamamura Kōzō, Japan Before Tokugawa.

were the means by which local areas were brought into the orbit of a regional power.

This had far less to do with conquest, and was primarily a process of elite endorsement and expansion of local practices which catalyzed progress towards peace.

Most people living in the Japanese archipelago during the late medieval era were only indirectly affected by battles between vassal warrior armies.24 Throughout the history of the islands, only about ten to fifteen percent of the population ever belonged to the ranks of a lord’s retainer band.25 From an estimated population of fifteen to seventeen million in 1600, that means probably somewhere around fourteen million people were not vassal warriors in late medieval Japan.26 Despite these demographic facts, a disproportionate amount of scholarly literature has focused on medieval warriors.27 Increased control of elite warriors over lower ranking ones is widely considered a key historical development of the late medieval period, and one with far- reaching implications. Not surprisingly, that process was protracted and contentious, and many studies have examined how rowdy medieval samurai were subdued by their social superiors. By analyzing disputes between upper and lower class warriors, various scholars have presented insightful conclusions. Major English-language works

24 Of course, many were directly affected, and this is not to suggest that elite warfare was somehow unimportant. But we must recognize the limited scope of samurai warfare. This was only one type of (armed) conflict within the broader context of the civil war era. 25 William Wayne Farris, Heavenly Warriors: The Evolution of Japan’s Military, 500-1300 (Cambridge: Harvard University Press, 1992), 339–42. Because “samurai” came to be an official, legally defined status in the late sixteenth century, I primarily use the broader term “warrior” to describe the class of hereditary professional military men who comprised the principle ranks of civil war era armies. Prior to the official designation of samurai as a status group by Hideyoshi in 1590 virtually anyone with the aptitude and means could become a military professional. As it was, there was no clear distinction between such professional warriors and other occupational groups in late medieval society. As we will see, even non-professionals bore arms throughout the civil war period. 26 Farris, Japan’s Medieval Population, 5. 27 To be fair, a major reason for the emphasis upon warriors is the availability of evidence. This study, too, relies heavily upon sources created by elites and warriors.

on the topic come from Eiko Ikegami, Michael Birt, David Eason, and to some extent

David Spafford.28 Each of these authors examines the evolution of warrior society in general, and disputes in particular, from a specific angle. Ikegami primarily examines social and cultural change, Birt economic, and Eason legal. Spafford’s work, though not mainly focused upon disputes, deals with a changing geopolitical landscape that required a shift in how warrior lineages constructed identity.29 Without question, elite warriors eventually achieved firm dominance over their subordinates and early modern samurai society became extremely hierarchical, regulated, and largely free of serious challenges to established sociopolitical relationships. These developments undoubtedly carried implications for society outside of the warrior class. Nonetheless, they do not represent the most salient and influential conflict for the vast majority of late medieval people.

What mattered most to the great majority was a different kind of contest: disputes over the local environment. Access to and maintenance of resources deemed necessary for survival motivated people to fight. So fundamental were these contests to late medieval society that historian Fujiki Hisashi 藤木久志 has argued our understanding of the word “battle” (kassen 合戦) as a clash of armies in the Sengoku Period is in fact wrong. Contemporaries understood that it could refer to such an event, but the common implication was something else. It was more often used to describe a conflict with an

28 Ikegami, The Taming of the Samurai; Michael P. Birt, “Warring States: A Study of the Go-Hojo Daimyo and Domain, 1491-1590.” (PhD diss., Princeton University, 1983); Birt, “Samurai in Passage”; Eason, “The Culture of Disputes in Early Modern Japan, 1550-1700”; Spafford, A Sense of Place. 29 Spafford examines the Uesugi 上杉 and shows how identity of the lineage shifted from one defined by their association with a particular office to one of a localized territory they physically occupied.

outsider, however defined, over some key resource.30 These included the following general characteristics: 1) they were between two parties to a dispute directly, in other words, without mediation, 2) they were very localized, and local claimants aggressively asserted their claims 3) all socioeconomic classes, including “ordinary” people (non- warrior retainers), had armed force at their disposal, and 4) as a result of the previous three, these disputes often resulted in violence.

The stakes of war for most people involved local resources. At its core, then, civil war era conflict was not a grand geopolitical struggle. As suggested above, it seems a distortion to prioritize struggles for supremacy amongst warriors. Japan’s long civil war was not primarily a political contest, nor an elite one. When major political change did finally arrive in the late sixteenth century, the extent to which it was revolutionary remains debatable. This period of disorder resulted from an endemic, longstanding tension between the center and periphery over how local production was to be distributed across the sociopolitical hierarchy. Weak political bonds between elites and their subjects proved inadequate to deal with changes to local/regional circumstances.

In other words, problems related to environmental management were key causes of civil war instability, and had to be corrected for order to be restored. This was a conflict to reconfigure the balance between local control and elite prerogative. Prior to the civil war, there had never been a particularly stable system in place which could effectively distribute local production across the sociopolitical hierarchy without conflict. What historian William McNeill has called the “macroparasitic balance” had long been volatile

30 Rather than, say, a clash of two armies such as the 1575 Battle of Nagashino. Fujiki Hisashi, Sengoku no sahō, 14.

between local, regional, and central actors in the Japanese archipelago.31 The creation of the early modern in the late sixteenth century marks the first time that such a stable balance was achieved.

Based on a variety of criteria, it is not hard to see why late medieval Japan is routinely described as tumultuous. It is important to place this particular kind of dysfunction in proper context, however. The civil war era was embedded within a set of historical circumstances that shaped its beginning and end. Both of those moments featured their own kinds of dysfunction, they had their own endemic problems. It is important therefore pinpoint the kinds of change pertinent to this study. There are 1) the prevalence of organized violence 2) effective (meaning regularized and largely unchallenged) resolution to disputes over very contentious issues such as local environment 3) overall continuity in ruling institutions, and 4) the relationship between central and regional/local power. I turn now to a sketch of how these factors have changed through three major economic paradigms in premodern Japanese history.

Each one had its own strengths and weakness, but as we will see, the relationship between local, regional, and central political authority had always been deeply problematic prior to 1600.

31 William H. McNeill, Plagues and Peoples (New York: Anchor Press, 1976), 63, 101.

B. Evolution from the Tributary to Market to Commercial Economy

The five centuries from 1150-1650 witnessed major economic and demographic growth. Working from figures pioneered by Japanese scholars, Wayne Farris argues that the population of the archipelago roughly trebled from 5.5-6.3 million in 1150 to

15-17million by 1600.32 These 500 years constituted Japan’s medieval period and are characterized by weak or no central government vis-à-vis strong regional power. The underlying trend was that central elites essentially lost control of the provincial economy beginning in roughly the twelfth century. For the rest of the medieval period, central authority became less and less important and a new type of economy, driven primarily by provincial commerce emerged. After years of internecine conflict, regional warlords were in a position to harness the provincial market economy to serve a mercantilist state. This mercantile order prevailed from roughly the mid sixteenth century to the early seventeenth. At that point, the balance of power once again began to swing back in favor of central authorities. Lasting peace under the Tokugawa order released a wave of prosperity that led to a highly sophisticated, consumer-driven, commercial economy. These trends were firmly in place by 1650.

From prehistory to the mid seventh century, tribal chiefdoms, and later hereditary kings governing proto-states covered the archipelago. No central government existed and power was localized, or at best, regional. This period of Japan’s history is traditionally divided into the Jōmon, Yayoi, Kofun, and Asuka epochs. Things changed rather dramatically in the seventh century. The adoption of Chinese-style institutions

32 Farris, Japan’s Medieval Population, 5.

beginning with the Taika reform of 645 marks the genesis of a new type of polity in the archipelago. For the first time, a central authority existed in Japan. This style of government reached its zenith during the Nara period (710-794).33 The Nara state had pretensions to imperial rule, and invested copious resources to create physical manifestations of imperial power. The city of Nara 奈良 itself and the great temple

Tōdaiji 東大寺 within it are perhaps the most vivid examples. The state also engaged in major infrastructure projects, building roads to facilitate the flow of goods from the countryside to the capital. Amino Yoshihiko notes that nascent states with designs on imperial power often construct roads as a way to exhibit their power and bind outlying regions under their authority.34 The Nara state was certainly ambitious in this regard.

Center dominated periphery for the first time.

Chinese institutions were implemented over the course of about seventy years from

645-718. Known as the ritsuryō 律令 (law and practice) codes, they were based upon a tributary, agricultural economic system. Divided into sixty-six provinces, each region of

Japan owed a specified yearly tax to the central government. Each person over the age of six was supposed to be allotted a plot of land to support themselves and pay taxes.

Taxes were to be paid as a portion of the rice yield from this land. However, there were many areas, and in some cases entire provinces, which lacked arable land suitable for rice cultivation. Government officials made adjustments, assessing tribute tax in terms of its rice value, but not necessarily rice itself. While agricultural regions did send rice to the capital, other provinces sent sea products, iron implements, silk, timber, baskets,

33 Some scholars consider the Nara period to have ended in 784 when the capital was moved to Nagaoka 長岡. 34 Amino Yoshihiko, Rethinking, 50.

or a number of other goods. Whatever form provincial tribute took, it continued to be measured by its rice equivalent, even if no rice was grown in the region. Imposition of

Chinese codes was an attempt to apply an economic and political system that was ill suited to conditions on the archipelago. Attempts at a Chinese-style army turned out to be an even worse fit than the tributary tax system. This army was utterly crushed in each major engagement it faced.35 Central elites simply could not make the Chinese sociopolitical order work in Japan. Nonetheless, the tributary agricultural system remained the structural basis of central authority for many centuries.

As early as the tenth century, serious economic and political dysfunction in the

Chinese-derived imperial system became visible. Aristocrats in the capital, Kyoto, relied upon a variety of local agents to do the actual business of tribute collection and transshipment. These individuals were nominally court appointees who served for set periods of time. In practice, posts became dominated by local elites and converted into hereditary privilege. Distant central elites had relatively little leverage over such officials, and had to find other ways to secure revenues from provincial land. The solution was the establishment of “personal” Estates (shōen 荘園).36 In this system, property was commended to a high ranking person in Kyoto, officially becoming theirs.

It was exempt from “official” tribute, but subject to payments to the nominal owner.

These elite “owners” then employed local agents directly as retainers, who were responsible for forwarding a range of payments from the Estate. Estates were effective

35 Farris, Heavenly Warriors, chap. 3. Conscription did not work because the court simply lacked the resources to train and outfit a large body of troops. Conscripts were expected to furnish their own gear, and train on their own time. Many found ways to avoid military service altogether. 36 The nature of Estates is discussed at length in chapter 1.

in maintaining elite revenues to an extent, but still suffered from the structural weaknesses of the tributary system. Rulers in the capital were almost entirely dependent upon the cooperation of local officials. That group collectively gained more and more leverage as time passed.

Estates were essentially a stop-gap measure enacted to maintain the tributary economic order. Their limited success gradually gave way to a new paradigm, one based upon the explosion of regional markets. This new market economy characterized the centuries from roughly 1250 through the first half of the civil war until the establishment of territorial states in the mid sixteenth century. It coincided with a shift away from central dominance to an increasingly atomized distribution of political and economic power. And it marks the ascendancy of the warrior class over the traditional aristocratic class. The establishment of Estates and the rise of the warrior class were reactions to large-scale societal changes in which provinces acquired greater economic and political clout. Largely driven by new technology, new crops, increased productivity, and an influx of cash, a reordering of central-regional relations converged around the mid twelfth century. Population figures hint at the importance the developments listed above. After a long period of stagnation, population grew steadily from 1150 to 1280, then accelerated exponentially after that.37 Underlying all those new developments was increased contact with the Asian mainland. Often traveling on more reliable Chinese ships, medieval Buddhist monks traveled with greater frequency to the continent. They brought back new forms of Buddhism, in particular Zen, which had a tremendous cultural impact. Journeys to China had significant material impacts as well

37 Farris, Japan’s Medieval Population, 5.

such as the introduction of the water wheel, terracing, champa rice (akagome 赤米), tea, and tofu.

Probably the most important Chinese agricultural imports of the medieval period were the water wheel and terrace farming. It is hard to definitively measure the impact of these technologies, but it is clear that agricultural productivity grew significantly in the late twelfth and early thirteenth centuries. Water wheels led to much more efficient irrigation, a perennial challenge facing Japanese farmers. Terracing made it possible to convert lowland hills into paddy, and evidence suggests that this was one of the key methods islanders used to increase arable acreage in medieval Japan. More people farming more land more efficiently contributed to economic growth. Hard figures are difficult to come by, but it is safe to say that more wealth existed in provincial Japan than had ever been the case.

Like the water wheel and terracing, champa rice arrived from Southeast Asia via traveling monks as well. Champa rice, or “red rice” in Japanese, ripens much more quickly than native Japonica strains. Farmers could plant and harvest two champa crops per year, essentially doubling output. Champa has the additional advantages of being drought and blight resistant, as well as having lower nutritional requirements. The upside was huge – a fast-ripening rice plant that could produce two crops per year, which was also hardier, and could be grown on poor soil with less water. Cultivation of champa rice and the ensuing double-cropping it made possible dramatically increased available calories for average Japanese in the twelfth century.

While many people undoubtedly benefitted from increased calories, few had access to meat, and for those not living near the sea, fish could be difficult to acquire. Tofu, yet

another monk-transferred import into the archipelago, made protein more readily available in the medieval Japanese diet. A staple ingredient in the meals served at

Chinese monasteries, tofu had existed in China for several centuries. Monks brought the soybean based product back to their own temples in Japan, and from there the methods of tofu production and preparation spread across the archipelago.

Tea drinking was elevated to an art form in the late sixteenth century, but like champa rice and tofu, it too came to Japan from China around the twelfth century.

Chinese monks found green tea to be particularly useful during long hours of Buddhist meditation and devotion. Seated meditation (zazen 座禅) required monks to remain awake but unmoving for extended periods of time. Sleep was the enemy of enlightenment, and tea helped the faithful stay alert. So too did the head monk’s strikes across the back with a bamboo dowel to dozing acolytes, but that practice had much less effect upon the Japanese diet as did tea drinking. When Zen monks brought tea home from China, it quickly became a popular beverage. By the thirteenth century, one could buy a cup of tea for a copper coin at one of the many tea huts nestled along roads, in towns, and in villages. Tea was quite a convenient crop to grow in Japan. The tea bush thrives along hillsides, which are numerous in Japan, and unsuitable for most staple crops. Farmers converted previously unused slopes into tea fields, and could then sell the plant. As tea drinking spread through society it contributed to a general rise in health along with champa rice and tofu. Tea has diuretic and antibiotic properties when drunk regularly, and since it is prepared with boiling water is much cleaner than drinking directly from a stream or well.

Tea and other Chinese imports stimulated productivity and demographic growth.

The provincial economy grew in tandem. Another Chinese import, coins, lubricated the wheels of commerce. In the early medieval period (1150-1450) a huge amount of

Chinese cash flowed into the archipelago.38 This facilitated exchange and ultimately specialization of trades. Expansion of rural commerce presented challenges and opportunities for people across the socioeconomic spectrum. It provided a huge opportunity for elites, although they collectively failed to realize it. Unable to reconfigure the land-based tax structure towards a more market-based system of extraction, the fortunes of central rulers stagnated. Regional elites exploited economic growth much more effectively. Kyoto aristocrats gradually lost their economic base and saw their political power slowly erode generation after generation. For commoners in the countryside, cash crops, commodities, and coins provided an opportunity to sidestep traditional means of taxation and keep more of their surplus in their own pockets. A tributary system based upon collection of goods in kind proved completely unworkable in the market environment of early medieval Japan. More and more tax payments were being converted into cash at regional markets. But major economic reform did not come. The market system prevailed for most of the civil war period until a new kind of political elite found ways to adapt market productivity in new ways.

In the civil war period, the administration of local productivity was highly localized.

It was loosely connected by inter-regional exchange to only a handful of major commercial hubs. Through processes explored in the following chapter, regional

38 Ethan Isaac Segal, Coins, Trade, and the State: Economic Growth in Early Medieval Japan (Cambridge: Harvard University Press, 2011).

magnates were able to encourage the growth of commercial hubs within their spheres of influence. During the second half of the sixteenth century, more and more daimyo capitals grew into economically advanced urban centers. These cities attracted specialists, wholesalers, and long-distance traders, who gained special privileges by offering their services to the newly ascendant class of warlords across the archipelago.

Old tributary relationships were finally abolished once and for all, and in their place was a new set of rights and obligations that was based upon the daimyo’s government.

The assessment and distribution of local productivity operated on fundamentally different terms than it had prior to the civil war. This new order was predicated upon a discourse of kōgi 公儀 “public authority,” which the daimyo claimed by way of his ability to maintain stability within the realm. At the climax of the civil war age in 1580,

Japan was ruled by several dozen such warlords, each controlling an independent mercantilist state.

Those territorial states were gradually incorporated into a larger and larger coalition led by a single dominant warlord. First Oda Nobunaga, then Toyotomi

Hideyoshi, and finally Tokugawa Ieyasu 徳川家康 (1543-1616) occupied this position.

After Ieyasu’s final victory in 1600, the economic order evolved from the mercantilist fragmentation of the late medieval age to a new kind of integration and consumption of the early modern period. Peace stimulated economic growth and interregional trade as never before in premodern times. New urban centers grew rapidly. They became sites of consumption on a much larger scale than ever before. Long distance overland and sea routes were now reliable and profitable due to the huge market available in the now dozens of large cities spread across Japan. This contributed to further specialization of

resource extraction, production, and trades in local areas. Specialization stimulated another productivity boom probably not seen since the 1200s, and population growth followed suit. As the archipelago became more integrated economically, local production became more and more oriented towards the market – the urban in particular. Commodification affected virtually every kind of resource, crop, and craft product. The resulting early modern commercial economy also gave rise to new culture.

It was for the first time an era when the common people took center stage as the dominant cultural force.

The evolution from tributary-market-mercantilist-commercial represents four stages in the balance between central and regional power. At the beginning, central power dominated. The center of gravity shifted to the regional and grew ever more localized during the market stage, which covered the early medieval period and into the civil war. The mercantilist order can be considered the pinnacle of a regional system, however within each daimyo domain, power became more streamlined and centralized.

It was a combination of strong central (warlord) authority with a high degree local autonomy. That system stayed in place into the early modern, commercial economy, but functioned on a larger, archipelago-wide scale. In the chapters that follow, I examine how these major transitions proceeded on either side of the civil war period, from the perspective of center-regional relationships and local environmental management.

I. From a Warring State to Warring States: The Collapse of the

Estate System and the End of Stability in Fifteenth Century Kai

This chapter examines the major causes of the civil war. By tracing a macroeconomic shift from a tributary to market-style economy, we can see how changes “on the ground” in local areas upset what was already a precarious relationship between central and regional authorities. In Kai Province, underlying economic tension ruptured the sociopolitical order in the opening decades of the fifteenth century. From that point on, Kai descended into internecine, localized warfare typical of the Warring

States Era. Instability persisted for the better part of the next two centuries, eventually ending at the conclusion of the Takeda-Tokugawa war in 1582. While the regional experience throughout the civil war was broadly similar to that of other areas, this chapter highlights key difference in Kai. Attention to regional peculiarities helps illuminate causes and effects in this confusing historical era with greater clarity.

In the fifteenth century, the traditional sociopolitical order that had prevailed for roughly five centuries broke down across the Japanese archipelago. This order relied upon patronage networks between various kinds of elites in and around the capital city,

Kyoto, and lower status elites based in the sixty six provinces (see Map x.1). As such, the authority of central and provincial elites was, in theory, ultimately predicated upon status rankings as determined by the imperial institution. Even after the decline of the imperial institution, which has traditionally defined the medieval period, the medieval sociopolitical order remained, discursively at least, an emperor-based system. The Ōnin

War (Ōnin no ran 応仁の乱) of 1467-1477 is often considered a turning point when

both the imperial institution and the warrior government headed by the Ashikaga shogun declined to the point that neither could exert any influence in the provinces.1

The war ushered in a new historical era defined by the prevalence of civil war. The use of Ōnin is however a convenient archipelago-wide estimation for a large-scale process which unfolded differently across regions.

In Kai, the tipping point between a frayed sociopolitical order and civil war arrived several decades earlier. It was catalyzed by an event known as Uesugi Zenshū’s 上杉禅

秀 (d. 1417) Rebellion (Zenshū no ran 禅秀の乱) of 1416-1417.2 For the remainder of the fifteenth century and most of the sixteenth, civil war conditions prevailed throughout Kai Province. Characterized by nonfunctional institutions, atomization of power, intensification of parochial rivalry, increase in local autonomy, ownership/control of territory based upon occupation, and near constant violent clashes between organized, armed groups, the civil war era in Kai, as elsewhere, was a chaotic age. As a result, in Kai conditions defining the “Warring States” period prevailed from 1417 until 1582, the date of the last major conflict in the province. It is therefore difficult to apply traditional periodization uniformly, and doing so obscures actual historical processes. 3

1 See for example Mary Berry’s monograph on the Ōnin War in Kyoto, in which the author argues that both Kyoto aristocrats and the Muromachi bakufu lacked any functional power outside of the city. Mary Elizabeth Berry, The Culture of Civil War in Kyoto (Berkeley: University of California Press, 1994). 2 Aka Uesugi Ujinori 上杉氏憲. 3 Medieval Japanese history has long been broken down into three major epochs with a handful of sub periods. The main periods are: Kamakura (1185-1333), Muromachi (1336-1573), and Warring States (1467-1590). Many variations of these main divisions exist, as do multiple definitions of sub periods such as the Nanbokuchō era (1336-1392), the “unification” era (1568-1590), the Azuchi-Momoyama era (1573-1603) and others. For my purposes here, I consider the medieval period to be from circa 1150 to 1600. This spans the political ascendancy of the warrior class, manifest in the career of Taira no Kiyomori 平清盛 (1118-1181) to the Tokugawa victory at Sekigahara 関ケ原. Main subdivisions of

Two large-scale trends in the medieval age unleashed widespread disorder. The primary cause was a series of technological, demographic, and economic shifts that concentrated greater wealth outside of the capital.4 Overall, the archipelago experienced a significant boom in local productivity beginning in the twelfth century and accelerating after 1280.5 This had the result of concentrating more wealth in the hands of local people, in particular local elites. Those individuals saw their leverage and influence increase vis-à-vis their superiors in the capital. That led to the secondary cause of civil war instability: the collective failure of central elites to revise ossified institutions in order to deal with changing local circumstances and the new center- regional balance. Government, at both the central and regional levels, continued to be based upon an emperor-centered tributary sociopolitical order. This underlying framework endured even under the Kamakura and Muromachi warrior-led polities, and significantly shaped behavior. Most importantly, the framework locked rulers into a set of anachronistic policies and practices which became increasingly out of sync with real conditions across the archipelago as time progressed.

Effects of the two trends identified above created the defining characteristics of the late medieval period. Central elites’ reduced economic and political hold over provincial areas decreased their overall authority, but it also meant that they no longer had a vested interest in maintaining order and resolving problems which arose in these places. The heads of the patronage networks that defined political relationships across

the medieval period employed in this study are twofold. Early medieval (1150-1450) is effectively the Kamakura and Muromachi periods. Late medieval is also referred to as the Warring States or civil war era, and is approximated as 1450-1600. As noted above, the Kai-specific definition of the Warring States era is 1417-1582. Further periodization is defined in text when employed. 4 See the introduction for a sketch these changes. 5 Farris, Japan’s Medieval Population, 1.

the archipelago did not (or more accurately could not) care what happened in local areas anymore. This meant that locals themselves were largely freed of obligations to central elites, but also lost the advantages of being attached to a functional institutional structure should they need it. In other words, local residents at all socioeconomic levels had to find ways to effectively govern themselves. Major tasks that had to be handled included assertion of their claims to territory and resources, law enforcement, judicial affairs, defense, perhaps most significant, conflict resolution with outsiders.6 Not surprisingly, this process was unstable and often violent, but over the course of the civil war era forged a new sociopolitical order.

In the early medieval era, economic and political power was concentrated in the hands what was known collectively as the kenmon 権門 (lit. “power gate”).7 It is useful to classify medieval kenmon along two axes: status and office (see figure 1.a). The three major status groups were aristocrats, elite warriors, and religious institutions. An individual might belong to multiple status groups, and there were likewise a dizzying array of offices that an influential person might hold in various organizations. For the analysis here, it is useful to categorize medieval office into three types: Estate (shōen) officials, bakufu officials, and provincial officials.8 Whatever the case, for the early medieval governance to function, there had to be a cooperative relationship between

6 My definition of the term outsiders is discussed in text below. 7 The metaphor comes from the fact that anyone with wealth and prestige was likely to live in a residence that had a gate. See Mikael S. Adolphson, The Gates of Power: Monks, Courtiers, and Warriors in Premodern Japan (Honolulu: University of Hawai’i Press, 2000). 8 The main Estate officials considered here are the landlords (honke 本家 or ryōke 領家) and on-site managers (jitō 地頭). Bakufu 幕府 is a term for the warrior-led governments of the Kamakura, Muromachi, and Tokugawa eras. Literally “tent government,” it is a reflection of military power structure of these polities. Provincial office came in many types, the most important here being that of the Provincial governor (kokushi 国司) and local magistrates (bugyō 奉行).

central elites in the capital and their regional subordinates. Estates relied upon distribution of various rights (shiki 職) to local production divided between officials at different levels. The most pertinent points of this relationship were between Estate landlords (honke 本家, ryōke 領家), who were almost exclusively central elites residing in the capital, and their on-site managers called jitō 地頭 (“land steward”) who actually oversaw local administration of the Estate.9 The shogun’s government (bakufu) relied upon lord-retainer relationships within the warrior class. Most elite warriors resided in the capital and exercised control over local areas via their retainer subordinates. The most elite provincial warriors, known as shugo 守護 (“military governor”), were ostensibly retainers of the shogun, although this relationship was seldom smooth.

Provincial officials included those officers of the imperial institution that had been established by the ritsuryō codes (ritsuryō-sei 律令制) in the seventh century.10 In theory, these posts were only open to members of the aristocracy (kuge 公家) possessing sufficient official rank. However, warriors, monks and other non-aristocratic people achieved provincial office in the medieval period through various means. There was considerable overlap between all three of these power structures (Esate, bakufu, provincial). That is to say, individuals could simultaneously occupy stations within the

Estate system, the bakufu, or provincial government in almost any configuration.

9 Honke and ryōke are traditionally labeled “guarantor” and “proprietor” respectively. I discuss their roles and my reasons for placing them in the same category of “landlord” in further detail below. The sketch of the shōen-kokugaryō system provided here is admittedly an oversimplification in order to emphasize the main thrust of this project. Although I have singled out landlords and managers, in reality, Estate administration was a tangled web of relationships between many more actors which has proven difficult for scholars to fully analyze. 10 Ritsuryō-sei is a contemporary term applied to a series of institutional reforms undertaken by the Japanese polity from the seventh through ninth centuries. These institutions were based upon Chinese models and had an enduring impact upon premodern governmental discourse.

Status (determined largely by lineage or occupation)

Office provincial (determined by Estate official bakufu appointment) aristocrats warriors religious institutions

Figure 1.a: Status Groups and Office of Medieval kenmon

In broad terms, local areas were administered in one of two ways: as Estate land

(shōen) or as provincial land (kokugaryō). Although these two institutional systems differed, in terms of how they functioned locally, they were largely similar. The subject of medieval local administration, on Estates and provincial land alike, suffers from a dearth of historical materials. Scholars have been forced to rely upon the relatively plentiful sources collected from the Kinai region. Much of what is known of outlying provinces including Kai must be inferred. Nonetheless, there is longstanding scholarly consensus that roughly half of all land was held as Estate and half as provincial territory during the medieval period. Maps of known Kai Estates support that estimate.11 Some ninety-nine separate Estates have been definitively identified in the province, with varying degrees of detail.12

In the early medieval age, governing institutions usually functioned as designed, with central elites maintaining an active role in the administration of provincial

11 Takeuchi Rizō, ed., Shōen bunpuzu (Tokyo: Yoshikawa Kōbunkan, 1975), 64–69. 12 Kokuritsu rekishi minzoku hakubutsukan, ed., Nihon shōen shiryō (Tokyo: Yoshikawa Kōbunkan, 1998), 332–40.

territory.13 By the late medieval centuries, the trend had reversed and institutions rarely functioned. Behind that shift lay the fracturing of linkages between capital and countryside. In order to understand what went wrong in the late medieval centuries, it is useful to briefly examine how the early medieval system actually functioned in a handful of cases involving local resource disputes. Unfortunately, no thorough accounts of local conflict resolution from Kai during this era have survived. In order to characterize local administration as it relates to disputes in early medieval Japan, we must consider examples from better-documented regions of western Japan. It is very likely that disputes in eastern Japan, in Kai and other provinces, were handled similarly.

The following section contrasts the structure of local administration and conflict resolution in the early medieval epoch with that of the civil war era and afterward. I use characteristic examples drawn from western Japan and in Kai to make this comparison.

A. Paradigms of Dispute in Medieval Japan

Both the logic behind local resource dispute resolution and its execution changed drastically over the course of the medieval period. We can classify the modes of local administration and it accompanying conflict resolution mechanisms into three types: the Estate, autonomous, and territorial regimes. Each featured a set of characteristics that evolved largely chronologically, although there are continuities and regional variations that make strict periodization difficult, even within a single province such as

Kai. Nonetheless, in general the “Estate” mode of local administration corresponds to

13 This role varied, and at times central rules enjoyed more or less influence over provincial areas in the era from 1150-1450. However, in general it is fair to say that their role was much more robust in these centuries when compared to the years 1450-1600.

the early medieval centuries of 1150-1450. The “autonomous” kind of sociopolitical order prevailed for the majority of the civil war era from 1450 until about the mid sixteenth century when territorial warlords gained firm influence over their domains.

When that happened, it ushered in a new kind of local administration which was a combination of high local autonomy and (comparatively) vigorous territorial polities.

This system became the foundation of the early modern sociopolitical order and thus had an enduring impact upon society.14

Medieval Japan scholars have long discussed the phenomenon of “self-help” (jiriki kyūsai 自力救済, lit. “self-powered aid”) as the underlying principle governing dispute resolution.15 Self-help discourse certainly remained influential throughout the entire medieval period. Conflict resolution depended upon one’s own action, not only in advocating for a solution, but in its enforcement. We might think such a system would readily dissolve into a crude “might makes right” scenario, but the way in which this discourse operated in medieval Japan was quite complex. A crucial component was that disputant’s actions had to be legitimized by “proper” authority. If they were not, they could be justifiably challenged and nullified. Of course, that still relied upon a party taking initiative to offer such a challenge, but nonetheless, this mechanism meant that the essential “rules” of conflict resolution, even for very specific circumstances such as local resource disputes, were more or less agreed upon by a large swath of the

14 Chronologically, it can be estimated as roughly 1550 to 1630, when the Edo political order was firmly solidified. Local administration at the very end of the civil war era was structurally similar to what prevailed for the majority of the early modern period. Key differences between the sociopolitical order of the late sixteenth century and early seventeenth were in the relationships between very elite warriors. 15 See for example the work of Fujiki Hisashi, Owada Tetsuo, Shibatsuji Shunroku, Asao Naohiro, Sakai Kimi, among others. In English, Wayne Farris, Eiko Ikegami, and David Eason have discussed medieval self help in detail.

population. The concept was in fact crucial to the functioning of the pre-civil war order.

It was why a challenge to that order such as Taira no Masakado’s 平将門 (d. 940) rebellion in 939-40, or later Zenshū’s rebellion in Kai, could be considered criminal.16

Within the Estate system, the implications of self-help for local conflict resolution meant that disputing parties dealt directly with one another and for the most part worked out solutions on their own. But when that proved impossible, locals turned to their high-ranking patrons, most of whom were aristocrats, elite warriors, or religious institutions based in Kyoto.17 In these instances, resolution itself had little to do with the actual local situation and was instead based almost exclusively upon the relative status of the people involved and their patrons.18

Elite individuals might wield great influence, but overall, the power of the central polity in medieval Japan was never impressive. There are a handful of high points surrounded by pools of dysfunction and decay. The end of the fourteenth century marks a rather precipitous downturn. Nevertheless, even at this time, certain central authorities maintained an influence over their provincial areas. One such figure was and Estate landlord named Fushimi no Miya Sadafusa Shin’o 伏見宮貞成親王 (1372-

1456), who held Fushimi Estate 伏見荘 of Kawachi Province 河内国.19 Sadafusa was a

16 Karl Firday discusses how legitimation worked amongst the ruling elite in the early medieval era in Karl F. Friday, Samurai, Warfare and the State in Early Medieval Japan, Warfare and History (New York: Routledge, 2004), chap. 1; For Masakado's Rebellion, see Karl F. Friday, The First Samurai: The Life and Legend of the Warrior Rebel Taira Masakado (Hoboken, N.J: John Wiley & Sons, 2008). 17 Tashiro Hiroshi provides a succinct overview of conflict resolution in the Estate system paradigm in Tashiro Hiroshi, “Chūsei kōki no shōen sonraku ni okeru funsō kaiketsu to ryōshu,” Kumamoto Shigaku 83–84 (2004): 79–108. 18 Ibid., 103. 19 Sadafusa is also known as Gosukō-in 後崇光院 and was Emperor Go-Hanazō’s 後花園 (1428-1464) father.

bit of an anomaly in his day because he actually lived on his Estate, and was of extremely high status as a member of the royal family. He did so, perhaps, because

Fushimi Estate was so close to the capital, located just south of the city proper where the Katsura 桂川 and Kamo 鴨川 rivers meet.20 In the fifteenth century, Fushimi Estate consisted of six villages, and Sadafusa took a rather active role in the administration of these communities in his role as Estate lord. He kept a journal spanning 1416-1448 known as Kanmon nikki 看聞日記 which provides a glimpse into a number of local resource disputes at Fushimi and how they proceeded under the Estate system mode of conflict resolution.21

Representative of Fushimi disputes is a conflict in 1420 between Fushimi Estate and neighboring Kowata Estate 木幡荘 to the east.22 It was a disagreement over grass clippings (kusa-kiri 草刈), most likely to be used as fertilizer. Kowata residents had cut grass from an area that Fushimi’s inhabitants considered their own. As the cutting proceeded, Fushimi people gathered in force and headed out to the disputed site. There they confiscated tools from the Kowata workers as a means to halt what was to them theft of an important resource. Both sides then prepped for an armed encounter, gathering “bows and arrows” (yumiya 弓矢). Before it came to that, however, Fushimi landlord Sadafusa learned of the situation and was able to get both sides to desist. He

20 Today Fushimi Ku 伏見区 of Kyoto. 21 Gosukō-in, Kanmon nikki (Tokyo: Kunaichō Shoryōbu, 2002). 22 The following details are summarized from Tashiro, “Chūsei kōki no shōen,” 86.

submitted a suit to the shogun and ultimately the dispute was resolved in bakufu court.23

Later that same year (1420) Fushimi avoided a potential feud with Higashi Kujō no shō 東九条荘 over water use thanks to their highly-placed patron.24 The dispute started after Fushimi residents appeals to Sadafusa that making their annual tribute payment

(nengu 年貢) would be difficult because of insufficient rainfall. They expressed a desire to pull additional water from the Katsura River, which flowed north to south through

Estate lands. Sadafusa responded by asking the landlord of Higashi Kujō, the Estate directly upstream from Fushimi, if his people could build additional sluices off the river in Higashi Kujō’s area. The lord, Kujō Mitsunori 九条満教 (1394-1449), agreed and apparently the issue was resolved without incident.25

Sadafusa and another elite Estate landlord used their influence to avoid serious violence once again in 1433.26 The trigger of this dispute is unknown, but a fight erupted near Ishida between people from Fushimi and five residents of Sumiyama

Village 炭山郷 as the latter were traveling to Kyoto. Fushimi residents injured three of the travelers, and took two of them hostage.27 Word somehow reached Sumiyama, probably via Fushimi, but in response a group of Sumiyama people set out and managed to grab two hostages from Fushimi. Fushimi countered by taking the other three individuals from the original Sumiyama group as they returned from the capital. This

23 To what result is unknown. Kanmon nikki reports that Sadafusa appealed directly to the shogun (kubō 公方), which may have been possible given his elevated status. The shogun at the time was Ashikaga Yoshimochi 足利義持 (1386-1428). 24 Tashiro, “Chūsei kōki no shōen,” 87–88. 25 Mitsunori is also known as Mitsuie 満家, and was the shogun’s regent (kanpaku 関白) at the time. 26 Tashiro Hiroshi 田代博志, “Chūsei kōki no shōen,” 85–86. 27 The other three were allowed to continue on to Kyoto.

exchange might have continued for some time and was almost sure to escalate. But the matter was resolved within two days when Sadafusa ordered all hostages returned after negotiating with Sumiyama landlord Sanbō-in Mansai 三宝院満済 (1378-1435).28

In all of these cases, we see that locals relied upon their high status patrons in certain instances. Elites had the ability to step in when circumstances grew violent or potentially violent. Of particular note here is that in all of these instances, no local agents of Sadafusa or the other Estate landlords were sent to investigate the situation

“on the ground.” Instead, Sadafusa advocated for his Estate by negotiating arrangements with other landlords. Though we do not know the details of these encounters, it seems very likely that Sadafusa leveraged his high status to achieve overall favorable terms in each of these cases. These can be considered typical aspects of local dispute resolution mechanisms within the Estate system. They can be clearly contrasted with examples from the “autonomous” system of the civil war age recounted next.

Not long after Sadafusa’s interventions on behalf of Fushimi, a very different kind of dispute system prevailed between Ōura 大浦 and Sugaura 菅浦 villages Ōmi Province

近江国.29 In 1445, these two villages clashed over the cutting of timbers from a contested area. When residents from Ōura cut wood from this place, the local militia

(wakshū 若衆) belonging to Sugaura gathered some twenty individuals and headed out to investigate. Both villages apparently felt entitled to the timber from this place. Ōura sent their own militia, and also requested assistance from a neighboring, allied village.

28 The details of these negotiations are not recorded in the journal. 29 The account of this dispute is summarized from Fujiki, Sengoku no sahō, 14–17.

The armed citizens failed to come to terms, and this incident ultimately resulted in a pitched battle. At the same time, a fourth nearby village sent their own militia to investigate, as the outcome of this fight undoubtedly carried implications for their own claims and interactions with neighbors. Armed residents of four separate communities thus converged on this occasion, with Sugaura squaring off against Ōura and its allies.

The Sugaura militia lost four people in the battle. Afterwards they reported to the village council (otonashū 大人衆) and advocated for a counterstrike on Ōura. But the council had decided to open negotiations with Ōura and the dispute was resolved without further bloodshed.

In this conflict, there was no high status patron available to intervene. Villages were fully on their own in terms of resolving problems, and in the civil war-era version of

“self-help” had to field their own militias in order to assert and protect their interests.

They dealt directly with rival parties, and as we see above, relied upon military force to achieve their goals. There were no external mechanisms available, no Sadafusa-type figure looming overhead who could pull strings and help solve local problems. The result was a propensity towards violence and a highly unstable system of dispute resolution throughout the civil war era.

Examples of the above trend can be found across all regions of the archipelago around roughly the same time. Local residents often resorted to violence in dealing with competing claims. Although examples of such action exist in the early medieval age, unsanctioned violence was proscribed and incurred punishment. Thus, conflict resolution during the civil war period was fundamentally different. It was localized, and meted out between autonomous groups on their own. Disputes over environmental

management proved to be some of the most contentious and persistent problems that local people faced. As provincial residents moved towards greater self-sufficiency, central elites grew increasingly impotent. Rivalries within the ruling class over an ever- shrinking portion of local production further contributed to the atomization of political power across the archipelago. We can see this process unfold in fifteenth century Kai.

B. Tensions within Center-Provincial Political Economy

Early medieval Kai was largely ruled by Kyoto temples, the Takeda lineage, and several dozen other warrior lineages of lesser status. Whether individuals resided on

Estate or provincial land, they were connected to one of these high-status landlords in largely similar ways. Some ninety-nine Estates can be confirmed for medieval Kai, but documents pertaining to proprietorship survive for only thirty-four of them.30 These landlords run the gamut from mid-level, regionally influential religious institutions such as the Hachiman Shrine 八幡宮 at Fuefuki 笛吹, to retired emperors via the Reizō-in 冷

泉院 holdings. Some of the higher status religious landlords included Ise Shrine 伊勢神,

宮 Ninnaji 仁和寺, Hōrinji 法輪寺, Jōkoji 淨居寺, Erinji 恵林寺, Shokokuji 相国寺, and

Kōshōji 光勝寺. Estate landlords held the lion’s share of economic and political power in the province until roughly the mid fourteenth century. They were also the group most likely to have direct control over local areas. Proprietors wielded authority on their Estates via on-site managers. Warriors were particularly well-suited to the tasks

30 Nihon shōen shiryō, 332–40.

of managing Estates for absentee proprietors, although monks occasionally found employment as managers as well.31

After the creation of the post of military governor by the first Kamakura shogun

Minamoto Yoritomo 源頼朝 (1147-1199), the top warrior in each province occupied this office. In Kai the post belonged to the Takeda from the thirteenth through fifteenth centuries, who ultimately parleyed their position into an autonomous domain as regional warlords during the civil war. The military governor served as a direct retainer of the shogun, and exercised power in each province via his own retainers. Local agents of the military governor did not necessarily owe allegiance to the shogun, an arrangement which led to frequent problems for the bakufu. In theory, the military governor’s jurisdiction extended to all provincial lands (i.e. kokugaryō 国衙領 – areas not held in Estate by a landlord), as well as to warriors who were employed by

Estate landlords. A military governor’s responsibilities included settling suits involving warriors, rendering military service to the shogun, and raising and commanding provincial troops for the suppression of rebellion or capture of criminals.32 Lesser warrior houses that served the Takeda or at times competed with them included the

Henmi 逸見, Uesugi 上杉, Atobe 跡部, and a branch of the Takeda known as Ogasawara

小笠原.33

In addition to Estate landlords and the military governor, there was a third type of elite provincial post, but it effectively functioned as another kind of Estate landlord.

31 Amino, Rethinking Japanese History, 71. Warriors were employed because of their skill set as policemen, guards, and fighters, while monks generally possessed knowledge of writing and mathematics. 32 Friday, Samurai, Warfare, and the State, 37. 33 Sasamoto, Sengoku daimyō, 105–129.

This was the office of the provincial governor (kokushi 国司), an old imperial post.

Provincial governors at one time played important roles in the administration of Japan’s provinces in the Nara period (710-784), but the functional importance of the office was short-lived.34 By the eleventh century nearly all provincial governors, including Kai’s, remained in the capital and sent agents called mokudai 目代 (“caretaker substitute”) to oversee provincial government. Those who actually resided in their appointed provinces earned a special name (zuryō 受領) indicating their real occupation of territory. Governors ostensibly administered all non-Estate land, but in fact played second fiddle to military governors. It was essentially a non-functional status title in the early medieval age. The post was thoroughly subsumed under bakufu authority by the mid thirteenth century. From that point on, military governors ruled non-Estate land while provincial governors derived their income from personal Estates. In other words, provincial governors joined the ranks of Estate landlords beginning in the thirteenth century. A key distinction between religious and aristocratic landlords in Kai vis-à-vis governor-landlords of the 1200s was that the latter physically resided on their Estates in the province. Regardless, for all provinces, Kamakura and later Muromachi shoguns controlled appointments to provincial governorships. Although by law these posts could only be awarded to aristocrats, appointments went to warriors almost exclusively from the 1250s. Shoguns preferred to appoint their own retainers as governors. In many cases, warriors holding the provincial governorship also claimed, legitimately or through fabrication, an aristocratic pedigree.35

34 Farris, Heavenly Warriors, 50–52. 35 In premodern Japan, the veracity of a genealogy was far less important, or often unimportant vis-à- vis the function it performed.

We do not have detailed information on who occupied the Kai governorship from the eighth to fourteenth centuries. Absentee officials of the Fujiwara 藤原 lineage may have held the position during this time.36 In the 1300s, the Ōi 大井 family received appointment as Kai governor from the Muromachi shogun. The Ōi consisted of two branches, one based in Shinano Province 信濃国 who were an offshoot of the

Ogasawara, and another based in Kai itself who were a branch of the Takeda. Both of these Ōi lineages traced their ancestry back to the Seiwa Genji 清和源治, as did the

Takeda. In other words, they had a claim to aristocratic pedigree, however circuitous.

By midcentury, the governorship of Kai became an empty title.

By the thirteenth century, the provinces were essentially being managed as Estates, whether they were shōen or kokugaryō land. This mattered because unlike the ideal,

Chinese-style imperial realm, Estates were attached to specific elite individuals or institutions. In other words, administration of territory was fractured amongst the ruling class. Estates originally provided personal income for Kyoto aristocrats and religious institutions. These landholdings were granted under the auspices of imperial authority, and existed as a portfolio of tax-exempt plots. There were two types of Estate landlords: the guarantor and proprietor. The guarantor was the highest ranking Estate official and the landlord his administrative agent. Originally every Estate was held by a proprietor and had to be “guaranteed” by a member of the imperial family or one of the most prestigious Buddhist temples. Explicit approval by the highest political authorities legitimized an Estate’s special tax-exempt status. Shortly after Estates appeared, laws

36 Yamanashi-ken shihensan iinkai, ed., Yamanashi kenshi tsūshi-hen 2: chūsei (Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 2006), 3.

requiring imperial guarantors were not strictly upheld. Lower-ranking aristocrats and temples began to occupy the post of guarantor. By the twelfth century, many Estates did not even have a guarantor. In these cases proprietors acted in the same capacity as before. Imperial family members and elite temples continued to use the title

“guarantor” for their own Estates, even if there was no proprietor, which was typical.

Whether an Estate belonged to a guarantor or proprietor, they were functionally the same and should be considered the landlord of the Estate in question. In rare cases past the 1100s when the letter of the law was fallowed and both a proprietor and guarantor were present, one of the two actually controlled the Estate and received its revenues.37

The other’s attachment to the Estate was in name only. Sometimes one forcibly removed the other.

Landlords did not own the land per se, but they were entitled to a percentage of its productive capacity. Estate managers and residents themselves each held specified rights (shiki) to portions of Estate products. Income came to landlords primarily in the form of the annual tribute (nengu), ideally a portion of the rice harvest. Nengu was effectively a levy upon the land, specifically agricultural land, of an Estate. Tenants also owed specified resources, goods, and services to the Estate lord. Tribute of non-rice products, as well as specialized services, was known as kuji 公事, literally “public work,” and translated here as duties. Corvee labor, called buyaku 夫役 (lit. “man-service”) existed as a separate category of unskilled work. Duties and corvee together were referred to as zōkuji 雑公事 “miscellaneous public work.” In contrast to the rice tax,

37 John Whitney Hall, “Terms and Concepts in Japanese Medieval History: An Inquiry into the Problems of Translation,” Journal of Japanese Studies 9, no. 1 (Winter 1983): 25.

zōkuji can be thought of as levies placed upon the people of an Estate.38 The manager receive a portion of these levees and was responsible for forwarding incomes to the landlord.

In those cases where Estate land was largely or wholly non-agricultural, rather than rice tax, duties and corvee comprised the bulk of annual tribute. The people of Kai sent wood, minerals, horses, leather, tools, fabric, charcoal, lacquer and other goods in addition to rice.39 Estate residents also paid rent (kajishi 加地子) to on-site managers of the Estate, which amounted to a kind of property portfolio for the manager himself.

Throughout Japan, proprietors ordinarily hired warriors to managerial posts. Of the known on-site managers for Kai, all were warriors.40 Originally, rent comprised the bulk of managers’ income on eleventh and twelfth century Estates. Ishii Susumu, Amino

Yoshihiko, Wayne Farris, and others have estimated that these managers received approximately one-third of Estate revenues in western Japan, with the remainder belonging to elite landlords.41 The opposite trend prevailed in eastern Japan, where collectively warriors had long held greater political and economic clout. So we can estimate that landlords typically received only about one third of the revenues of Kai

Estates, while two thirds stayed in the hands of managers. This may help explain why local warrior lineages enjoyed such autonomy and influence in the region.

38 Jeffrey P. Mass, The Kamakura Bakufu: A Study in Documents (Stanford: Stanford University Press, 1976), 202. 39 Yamanashi kenshi tsūshi-hen 2, 104. 40 Nihon shōen shiryō, 332–40. 41 Farris, Heavenly Warriors, 344–349; Jeffrey P. Mass, The Development of Kamakura Rule, 1180- 1250: A History with Documents (Stanford, Calif.: Stanford University Press, 1979), 48–55; Abe Takeshi, Nihon shōen-shi (Tokyo: Ōhara Shinseisha, 1972), 206–211; Amino Yoshihiko, Higashi to nishi no kataru nihon no rekishi (Tokyo: Kōdansha, 1998), 207; Kawane Yoshihiro, Chūsei hōken-sei seiritsu shiron (Tokyo daigaku shuppankai, 1971), 121–152; Nishioka Toranosuke, Shōen-shi no kenkyū (Tokyo: Iwanami Shoten, 1956), 739; Inagaki Yasuhiko, Shōen no sekai (Tokyo daigaku shuppankai, 1973).

While landlords had no claims upon provincial land, they did draw income from certain types of non-Estate land. Artisans of various trades, including hunters, tanners, butchers, shaman, traveling performers, prostitutes, and ascetics lived in backwoods areas called sanjo 山所, literally “mountain places.” Their status was not as clear as individuals who resided in established settlements. Although backwoods were not part of an Estate, these specialized tradesmen owed goods or services to aristocratic patrons

(i.e. Estate landlords) in exchange for tax exemption. The exact nature of the outcast- landlord relationship is extremely difficult to determine, but nonetheless provided essential materials to central elites.42

What proprietors could not acquire directly from their Estates, or from associated artisans, they purchased at regional markets. Many Estate holders sold surplus from the rice harvest in order to buy necessities which could not be acquired directly from their lands. In medieval Kai (1185-1600), Estate-holding temples and shrines often hosted regional markets. Omuro Sengen Shrine 御室浅間神社, located in the northern foothills of Mt. Fuji, was actively engaged in commerce throughout the fifteenth and sixteenth centuries. It is clear that the shrine relied heavily upon trade for income as well as necessary goods. The chronicle Myōhōji-ki 妙法寺記, written by a Nichiren monk overseer of Sengen Shrine, regularly records periods of extreme “difficulty” (tsumaru

42 Unattached outcasts who provided specialized goods and services were initially organized into trade guilds under the patronage of the imperial family. This system was institutionalized in the ritsuryō codes of the seventh and eighth centuries, but its effective implementation has proven to be obscure. We do know that elites were able to acquire backwoods resources somehow. By the medieval period my guess is that most Kai elites did so via regional markets, while central aristocrats relied upon an emerging class of professional merchants. See Janet R. Goodwin, Selling Songs and Smiles: The Sex Trade in Heian and Kamakura Japan (Honolulu: University of Hawai’i Press, 2007); Amino, Rethinking Japanese History; Amino, Muen, kugai, raku.

koto kagiri-naku ツマル事無限) in Kai when buyers and sellers failed to patronize local markets.43

Although the ritsuryō codes stipulated a strong central government, a large bureaucracy staffed by ranked aristocrats, and an officialdom ultimately answerable to the emperor, provincial land during the Estate System was governed much like Estates themselves. One major different in Kai was that Takeda military governors actually resided in the province throughout their tenure as bakufu officials. The family was based at Ichikawa Estate 市川荘 in Koma County 巨摩郡 until Takeda Nobutora 武田信虎

(1493-1574) moved the headquarters to Kōfu, a settlement in the largest basin in the region, in 1519. Although they were bakufu officials, as mentioned above, the Takeda were also Estate landlords. That arrangement was not uncommon for other military governors and lessor bakufu officials. Some scholars have characterized this as a divide between the “private” administration of a warrior’s personal Estate versus his “public” administration of provincial land.44 These can be useful in understanding the complex

43 Fujiyoshida-shi shihensan-shitsu, ed., Myōhōji-ki, Fujiyoshida-shi shi shiryō sōsho (Fujiyoshida- shi: Fujiyoshida-shi kyōiku iinkai, 1991). This document hails from the Katsuyama area of Kai (Map 4.1) and spans the years 1466-1561. Little is known about its production, but at least a portion of it, from circa 1500 until 1528 was written by Nikkoku 日国 (d. 1528), abbot of Jōzaiji 常在寺, in the modern town of Fuji Kawaguchiko 富士河口湖町. From 1533 until the end of the chronicle, the author(s)’ and compiler(s)’ perspective is from Yoshida Village 吉田郷, not far from Lake Kawaguchi. These entries contain great detail about local affairs, including weather, crops, commerce, and conflicts between villagers and rural samurai. The document is considered a highly reliable historical source. A nearly identical chronicle called Katsuyama-ki 勝山記 also originated in this same area, and there is much debate amongst scholars over the relationship between these two versions. This paper primarily relies upon the 1826 transcription of Myōhōji-ki known as the Bunsei Kakkou-ban 文政活刻版, although I have consulted a second transcription and Katsuyama-ki as well. All three of these variations can be found in the volume cited here. 44 The eminent historian Jeffrey Mass was the first to argue for the significance of a public/private dichotomy in medieval Japan. Mass’s views were influential upon a number of subsequent scholars. See for example Jeffrey P. Mass, ed., The Origins of Japan’s Medieval World: Courtiers, Clerics, Warriors, and Peasants in the Fourteenth Century (Stanford: Stanford University Press, 1997); Karl F. Friday, Hired Swords: The Rise of Private Warrior Power in Early Japan (Stanford, Calif: Stanford University Press, 1992); Thomas Conlan, State of War: The Violent Order of Fourteenth-

roles military governors played. However, since military governors essentially ruled provincial lands in the same manner as Estates, the public/private dichotomy meant very little in practice. Rather, the military governor can be considered the landlord of his personal Estate, and the manager of provincial land. In other words, provincial land was similar to an Estate, but with the shogun as the central landlord and the military governor an on-site manager. In both cases, it is essential to realize that authority and rights flowed through personal, hierarchical patronage networks – in other words vassalage. Thus, the public and private distinction becomes somewhat artificial.

Provincial land was really no more public (or private) than a warrior’s own Estate; he just held different rights and responsibilities upon that land.45

C. Uesugi Zenshū’s Rebellion and Instability in Kai

The political and economic arrangements described in the preceding section proved to be very problematic. Numerous problems surfaced in the form of rebellions, regime change, and localized conflict throughout the early medieval period. In Kai, the established order ultimately unraveled for good in the early fifteenth century, ignited by a rivalry between the shogun and one of his chief officers in eastern Japan.

The second of Japan’s warrior led governments, the Muromachi bakufu, was the presiding central regime in the fifteenth century. This polity had faced serious challenges to its authority since its creation in 1336. While Ashikaga shoguns controlled

Century Japan (Ann Arbor: Center for Japanese Studies, University of Michigan, 2003); Spafford, A Sense of Place. My own views are more in line with that of Wayne Farris and Mikael Adolphson, who consider the public/private split a misleading portrayal of medieval political discourse. See Farris, Heavenly Warriors; Adolphson, The Gates of Power. 45 The basic political and economic relationships described in this section are represented graphically in section 1 of the appendix (figures 1.b-d).

Kyoto and the imperial court, for almost sixty years from 1336-1392 a rival court based at Yoshino 吉野 in Yamato Province 大和国 (modern Nara Prefecture 奈良県) fought against the bakufu. This is known as the Period of Northern and Southern Courts, or

Nanbokuchō 南北朝 Era. Muromachi armies finally eliminated the Southern Court in

1392, but that did not end the bakufu’s troubles. The shogun had always had difficulty controlling the eight provinces of the Kantō 関東 region, and the far northeast of

Honshū, Dewa 出羽 and Mustu 陸奥 provinces. Ashikaga Yoshimitsu 足利義満 (1358-

1408), shogun from 1367-1394, came up with a solution. He appointed a deputy, to be named from the Ashikaga house, as governor general, Kamakura kubō 鎌倉公方 (lit.

“lord of Kamakura”), who would rule the troublesome Kantō provinces on the shogun’s behalf.

Based in the old shogunal capital at Kamakura, the governor general acted with the shogun’s authority, and answered to him alone. Despite nominal subjugation to the bakufu, the Lord of Kamakura wielded near unilateral power in eastern Japan. It was almost as if there were two warrior governments, one in the west headed by the shogun in Kyoto, and one in the east based at Kamakura. The Muromachi period east-west split was not unlike that of the earlier Kamakura period diarchy.46 A key difference, of course, was that the Kamakura diarchy consisted of an aristocratic government in the west, and a warrior government in the east. Parallels aside, various powerful warrior lineages served the governor general. Chief among these was the eastern deputy, Kantō kanrei 関東管領, the lord of Kamakura’s second in command. These retainers of the

46 Wayne Farris and Ethan Segal have discussed the concept of diarchy and its applicability to early Kamakura era Japan. See Farris, Heavenly Warriors, 1992; Segal, Coins, Trade, and the State.

governor general did not owe loyalty directly to the shogun, an arrangement that rarely proceeded smoothly.

In the early fifteenth century, relations between the shogun and the governor general progressed from dysfunction to open conflict. A shogunal retainer called Uesugi

Zenshū, also known as Ujinori 氏憲 played a key role in the derioration of relations between Kamakura and Kyoto. In 1411, Zenshū became eastern deputy (kantō kanrei) to Governor General Ashikaga Mochiuji 足利持氏 (1398-1439).47 Zenshū belonged to the Inukake 犬懸 branch of the Uesugi, one of four branches from which governors general traditionally chose their eastern deputies.48 Mochiuji appointed Zenshū to the post at the behest of his uncle Ashikaga Mitsutaka 足利満隆 (1392-1417), a strong

Inukake patron. Mochiuji himself favored the Yamanouchi 山内 branch, and never got along well with Zenshū. The battle chronicle Kamakura ōzōji 鎌倉大草紙, one of the main sources for Zenshū’s Rebellion, dwells upon animosity between Zenshū and

Mochiuji at length.49

Despite the enmity between the two, Zenshū assumed the post of eastern deputy, making him one of the most powerful warriors in the east. The military governor of Kai,

Takeda Nobumitsu 武田信満 (d. 1417), married Zenshū’s daughter in the spring of

1413, solidifying Takeda prestige in the region. Tying Takeda fate to the brash Zenshū

47 Ujinori was Zenshū’s given name. He took the Buddhist name Zenshū in 1415. 48 The other Uesugi branches were the Yamanouchi 山内, Takuma 宅間, and Ōgigayatsū 扇谷. 49 Yamanashi kenshi tsūshi-hen 2, 222. This document is often considered a sequel to the Taiheiki 太平記, the chronicle of the Ashikaga bakufu. Authorship is unknown, but could possibly have been by the monks Tō Tsuneyori 東常縁 (d. 1484) or Saitō Myōchin 斎藤妙椿 (1411-1480). It was an heirloom of the Chiba 千葉 family.

proved to be fatal for Nobumitsu. He was more or less dragged into Zenshū’s rebellion, ultimately at the cost of his life and much of his family’s influence in Kai.

Tensions between Zenshū and Mochiuji flared in 1415 when the shogun confiscated the fief belonging to an Uesugi vassal named Kobata Rokurō 越幡六郎 (unknown).50

The nature of Kobata’s violation was unspecified, but Zenshū considered it unjust and he angrily demanded his vassal’s reinstatement. Mochiuji declined in no uncertain terms. In protest, Zenshū’s refused to perform his service to the bakufu, which then prompted Mochiuji to strip of his office as eastern deputy. In Zenshū’s place was appointed Uesugi Norimoto 上杉憲基 (1392-1418), of the Yamanouchi branch,

Mochiuj’s close allies. According to the Kitsuregawa 喜連川 family genealogy, this whole series of events begun by the punishment of Zenshū’s retainer Kobata had been a ploy to get the Yamanouchi back to the post of eastern deputy.51 Zenshū apparently arrived at the same conclusion, and soon afterwards made his move against Mochiuji.

Inukake Uesugi aside, Mochiuji had enemies in the capital as well. Taking advantage of the brewing hostility in the east, Mochiuji’s half brother Ashikaga Yoshitsugu 足利義

嗣 (1394-1418) conspired with Zenshū against the unpopular governor general.

Yoshitsugu employed a traveling Zen monk to coordinate plans with an emerging anti-

Mochiuji coalition in Kyoto and Kamakura. At the appointed time, Mochiuji’s uncle

Mitsutaka and his nephew (Mochiuji’s cousin) Mochinaka 持仲 (d. 1417) traveled to

Kamakura, where they joined with Zenshū and his warriors, including Takeda

50 Ibid. 51 Ibid. The Kitsuregawa were a minor branch of the Ashikaga, tracing their lineage from Mochiuji.

Nobumitsu. The rebels then stormed Mochiuji’s mansion on the second day of the tenth month of 1416.52

Zenshū’s attack did not succeed in killing the governor general, but it did force

Mochiuji to flee to Suruga Province to the west. There, he took up residence with

Military Governor Imagawa Norimasa 今川範政 (1364-1433), while in Kamakura,

Zenshū reclaimed his old post as eastern deputy and Mitsutaka appointed himself governor general. It was to be a brief reign for the co-conspirators. Shogun Ashikaga

Yoshimochi (1386-1428) issued urgent orders to military governors of the kantō to dispatch the rebels. Fortunately for the bakufu, these eastern warriors still saw more opportunity for advancement in service to the shogun rather than with the upstart

Zenshū faction. A combined force of warriors from various eastern provinces fell upon the rebels at Seyahara 瀬谷原 (modern Yokohama) and crushed them on the ninth day of the first month of 1417.53 The very next day, Zenshū, Mitsutaka, and Mochinaka performed ritual suicide (seppuku 切腹) at the shogun’s villa Yukinoshita 雪ノ下 in

Kyoto.54 Takeda Nobumitsu had either escaped Seyahara, or had not been present. In either case, the shogun quickly ordered his arrest and charged Uesugi Norimune 上杉憲

宗 of Echigo Province to do so.55 Uesugi’s warriors pursued Nobumitsu to Tokusayama

木賊山 in Kai’s Tsuru County 都留郡 and killed him in battle.56 After the death of Kai’s

52 Ōdai-ki in Hattori Harunori and Shimizu Shigeo, eds., Takeda shiryōshū (Tokyo: Jinbutsu Ōraisha, 1967), 346. 53 Ibid., 347. 54 Ibid. 55 Yamanashi kenshi tsūshi-hen 2, 223 Norimune held the governorship of Awaji Province 淡路国, but likely had never even been there. 56 Subsequently known as Tenmokuzan 天目山, this was the same site where family head Takeda Katsuyori 武田勝頼 (1546-1582) met his demise several generations later.

military governor, Kamakura ōzōji reported that “all the commoners were in rebellion” throughout the province.57 Chaos ensued for the next two decades.

Eager to quell the power vacuum in Kai, the shogun named Takeda Nobumoto 武田

信元 (unknown) military governor.58 This was a reasonable choice. Nobumoto possessed Takeda pedigree, the family long associated with this post, and he could ultimately claim descent from the Kai Genji. Mochiuji, now back as governor general after his victory over Zenshū and cohort, had his own candidates in mind for Kai military governor. Having weathered a serious challenge to his authority, and widely regarded as over-ambitious, Mochiuji’s posturing raised suspicions in the capital.

Shogun Yoshimochi was interested in checking Mochiuji’s growing influence, and so had much riding upon the success of his chosen appointee Nobumoto in Kai. Unfortunately for him, Nobumoto was not up to the job of restoring order in the region. Nobumoto had been living as a monk at the Mt. Kōya 高野山 temple complex for over a decade. He did not enthusiastically return to secular life, but rather had to be escorted to his new post by his cousin Ogasawara Masayasu 小笠原政康 (1376-1442), military governor of

Shinano. The pair arrived with their retinue in the second month of 1418, over a year after Zenshū’s rebellion, and immediately had to battle the Henmi family. Henmi warriors allied with Mochiuji during the rebellion, and their leader Arinao 有直

(unknown) had designs on the military governorship.

57 Yamanashi kenshi tsūshi-hen 2, 225. 58 Ibid., 226 Nobumoto came from the Anayama 穴山 family, a Takeda branch that had split generations ago. He was known as Anayama Mitsuharu 穴山満春 before becoming Kai military governor in 1417.

Nobumoto was not an effective military governor. He did his best to restore order to the province and stay alive, but achieved neither. Nobumoto’s death in 1420 created another struggle between the governor general and the shogun.59 Neither side could agree upon the next appointment to Kai military governor. Mochiuji claimed it was his prerogative as governor general. The shogun Yoshimochi countered by saying that

Mochiuji had forfeit his powers of appointment since he had been the one to name the traitor (in the eyes of the bakufu) Takeda Nobumitsu, who had fought with Uesugi

Zenshū, to the post years earlier.60 As these two continued to bicker, the Henmi expanded their influence in Kai. Finally, the shogun settled on Takeda Nobushige 武田

信重 (1386-1450), Nobumitsu’s son, in 1421.

Nobushige’s tenure as military governor vividly illustrates two characteristics of early fifteenth century political authority in Kai. The province was out of control, and the bakufu could do nothing about it. In fact, it was so dangerous that Nobushige refused to actually take up his post in Kai. Instead, he remained in the capital and appointed Henmi Arinao, the old Takeda rival, as his deputy. Arinao thus became the de facto military governor, but his own grip on the levers of provincial government was severely contested. Arinao was in fact openly battling a number of rival warrior lineages. Meanwhile, Yoshimochi repeatedly ordered Nobushige to travel to Kai and assume his responsibilities. Nobushige displayed more interest in self-preservation

59 The circumstances of Nobumoto’s death are unclear. He likely died in battle against the Henmi. See Sengoku jinmei jiten iinkai, ed., Sengoku jinmei jiten (Tokyo: Yoshikawa Kobunkan, 2006), 397. 60 Of course, Zenshū and his allies had actually rebelled against Mochiuji, not the shogun. Nonetheless, Yoshimochi used this as justification that Mochiuji’s judgment was suspect and that he should not be the one to select the next military governor of Kai. His reasoning was immediately undercut when he ultimate chose Nobumitsu’s son Nobushige for the position. What this shows is that the back and forth between Mochiuji and Yoshimochi had nothing to do with naming a qualified person as Kai military governor but was a power struggle between the two of them.

than prestige, and candidly refused to go. Indicating the deterioration of shogunal authority by the fifteenth century, Yoshimochi was unable to force his own military governor from the capital.

As it became clear that Nobushige would not be coming to Kai to be the military governor, competition between warrior lineages in the province intensified.

Nobushige’s brother Takeda Nobunaga 武田信長 (d. 1477) was determined to eject the upstart Henmi lineage and reassert Takeda supremacy in Kai, with himself as family head. He rallied retainers and old Takeda allies to oust the deputy Arinao and take his brother’s place as military governor. Gradually, nearly all warrior families in Kai were drawn into a Takeda vs. Henmi war. They split into two camps, the League of the Sun (hi ikki 日一揆) backing Nobunaga, and the Ring of Wealth League (rinpō ikki 輪宝一揆) supporting Arinao. These two factions fought for over a decade with little progress.

As warfare raged on in Kai, Mochiuji’s continued to clash with the shogun, even after

Yoshimochi’s death in 1428. The ever ambitious governor general became infuriated when the office of shogun went to his uncle Ashikaga Yoshinori 足利義教 (1394-1441), a son of Yoshimitsu. After several years of simmering tension, open hostility erupted in

1432.61 Mochiuji decided to intervene in the ongoing Takeda-Henmi war in Kai, ostensibly to restore order. Contemporaries knew, however, that Mochiuji’s true intent was twofold: 1) settle old grudges against the Takeda, and 2) to get his own agents installed to the ruling posts in Kai. The bakufu ordered Mochiuji to stay out of Kai, probably because the shogun was worried that successful intervention would boost

61 Yamanashi kenshi tsūshi-hen 2, 235.

Mochiuji’s influence across the east. However, the shogun could do little to actually control Mochiuji.

Since Yoshinori was relatively impotent in preventing Mochiuji from getting involved in the fight in Kai, he retaliated via insult. In 1438 Yoshinori declined to bestow a name character upon Mochiuji’s son, a very public slight and a serious one at that. Governor General Mochiuji then turned his warriors against the shogun.62 For him it ended up being a fatal mistake. Yoshinori mustered enough of an army to finally corral Mochiuji at Eianji 永安寺 in Kamakura. Although Mochiuji appealed for mercy in exchange for taking the tonsure, the shogun refused. In response, Mochiuji ritually killed himself on the temple grounds rather than undergoing capture.63

Scholars point to the Ōnin War as the deathblow that ended the Muromachi sociopolitical order. It had ended much earlier in Kai, however. By the outbreak of Ōnin,

Kai had been embroiled in conflict nearly continuously for fifty one years. Central rule had been effectively powerless since the aftermath of Zenshū’s Rebellion. Since then, autonomous villages, local warriors, and ultimately regional warlords ruled Kai independent of any central authority. For the next century, regular violence erupted between these three entities. Eventually, Takeda warlords reestablished a modicum of stability, solidifying their rule around 1520. But, multiple kenmon continued to exist at the local level and fought for supremacy. It took another sixty years or so to completely pacify the province. Civil war conditions ultimately ended in Kai with the Tokugawa takeover in 1582. Thus, “Warring States” Kai can be best understood as 1417-1582.

62 This conflict is known as the Eikyō Rebellion 永享の乱. 63 Ōdai-ki, 347.

The political dysfunction which erupted in fifteenth century Kai and other provinces was a symptom of major changes in the local economy. Various structural instabilities made it impossible for the established sociopolitical hierarchy to adapt. Central elites watched nearly powerlessly as regional and local players claimed ever more autonomy.

Understanding why this happened is a thorny historical problem for which scholars have yet to arrive at a convincing conclusion. Although my treatment of it here has been brief, I am convinced that further investigation into the local economy and local practice will yield valuable insight. Moreover, we need a better a better sense of why social and political precedent remained so influential in premodern Japan. A robust cultural history of precedent is beyond the scope of this work. However, the closing section presents some general observations as they relate to the current topic.

D. Conclusion: Administrative Anachronism

Since the institution of a Chinese-style government in the seventh century, the economy of the Japanese archipelago was designed to function as a classic center- periphery tributary system. Administrative codes designated the capital as the site of large scale buying and selling. In the eyes of the aristocratic state, provincial commerce should have entailed no more than petty exchanges for daily necessities at local markets. But the early medieval proliferation of provincial markets, and indeed of

Estates themselves, were symptoms of a long-term economic shift which eroded the effectiveness of this tributary system. As Ethan Segal has shown, these markets were integral parts of the overall economy, and had little to do with the central government

from at least the tenth century onwards.64 A marked increase of agricultural surplus around 1250, higher concentration of wealth outside the capital, and an influx of coins combined to disrupt the economic system prescribed by the ritsuryō codes.65 What was designed to be a center-periphery tributary economy became a regionalized market system consisting of periphery-periphery exchanges in which the importance of the capital was greatly reduced.

We can think of this evolution in terms of socioeconomic status as well as geography. Aristocrats lived in Kyoto, for them the all important center of culture, legitimacy, and wealth. All but the most elite warriors made their living in the provinces. We have seen that many worked on Estates, served military governors, or found employment with the bakufu itself. Despite the clear shift of power away from aristocrats and towards warriors, both the Kamakura and Muromachi bakufu continued to derive legitimacy, rank structure, and administrative mechanisms from the framework of the aristocratic state. For that reason, in terms of governance, little changed structurally over the course of the Estate System. What did change was the level of real control exercised by the Kamakura and Muromachi shoguns’ governments.

As in Estate land, power shifted from central elites (predominantly aristocrats) to provincial elites (predominantly warriors) steadily from 1000-1450 and became concentrated solely in local hands after the mid fifteenth century. A landmark edict of

1352 has been cited by scholars as a salient indication that central elites had lost their

64 Segal, Coins, Trade, and the State, 21. 65 Ibid.; Farris, Japan’s Medieval Population, chap. 4.

economic hold over provincial territory.66 The so-called “half tax” (hanzei 半済) issued by the bakufu granted military governors rights to fully one half of the total revenues of provinces under their purview. This was essentially an admission that the central government had no real oversight over the governors and is considered to be the genesis of the shugo daimyō 守護大名 phenomenon.67

Muromachi-era politics has often been characterized as a delicate balance of power between the shogun and the sixty to seventy military governors in each province.68 At times a vigorous shogun could wrangle these provincial warriors into submission, as during the reign of Yoshimitsu. Usually though, the shogun’s nominal subordinates collectively held the upper hand. In either case, the Muromachi government never held expansive authority. Evaluated from any standpoint, it cannot be considered a successful government. By the fifteenth century, Ashikaga leaders were lucky if they survived their tenure as shogun. Of the final ten Muromachi shoguns, from 1429-1473, nine were either assassinated, died in war against their own retainers, or died in exile.69

James Murdoch famously labeled the era “…the golden age, not merely of turncoats, but of mediocrities.”70 Contemporary scholars are not so harsh in their assessment of the

66 William Wayne Farris, Japan to 1600 : A Social and Economic History (Honolulu: University of Hawaii Press, 2009), 140. 67 Essentially, the military governors as autonomous rulers of provinces as a fief or principality and only loosely subject to the shogun’s authority. 68 Kozo Yamamura, The Cambridge History of Japan, vol. 3 (Cambridge: Cambridge University Press, 1990). 69 Only Ashikaga Yoshikatsu 足利義勝 (1434-1443) did not die in such fashion. He reigned as shogun for only one year, 1442-43, at age eight, and suffered a fatal fall from horseback. 70 James Murdoch, Isoh Yamagata, and Joseph H. Longford, A History of Japan, (Yokohama: Kelly & Walsh, 1903).

period, but nonetheless continue to emphasize the structural weakness of the

Muromachi system.71

The Muromachi government was so unsuccessful in large part because it was economically anachronistic. Increased productivity helped local people assert greater independence. Ruler’s lost their ability to enforce tributary relationships prescribed by the imperial state system, and from the bottom, no strong incentive to maintain those relationships remained. Nonetheless, warrior governments, eager for legitimacy and prestige, continued to operate within the traditional political discourse. The first two bakufu tried, and ultimately failed, to prop up an increasingly non-functional institutional framework. There was no drastic fiscal reform initiated by either

Kamakura or Muromachi leaders in order to deal with the economic changes which accelerated in the thirteenth century.72 As a result, the political system centered upon

Kyoto, the emperor, and extraction of tribute from the provinces, gradually moved more and more out of sync with reality. That is to say, elites transplanted institutions designed for a vigorous imperial state onto what had become became a conglomeration of regional markets and power centers loosely connected by trade and vassalage ties between elite warrior-administrators. Moreover, Segal argues that economic reforms initiated under strong Ashikaga shoguns such as Yoshitmitsu actually accelerated the shift of power from capital to province because they did not address structural problems.73 Instead of re-centering the capital at the top of the political and economic

71 Representative examples include H. Paul Varley, The Ōnin War; History of Its Origins and Background with a Selective Translation of the Chronicle of Ōnin (New York: Columbia University Press, 1967); Berry, The Culture of Civil War in Kyoto; Conlan, State of War. 72 Segal, Coins, Trade, and the State, 65. 73 Ibid., 184.

hierarchy, Yoshimitsu’s reforms amounted to a series of concessions to regional power in exchange for increased cooperation over the short term. The bakufu of the 1400s thus became a government designed to administer an economy that no longer existed.

By that time, the reality across much of the Japanese archipelago resembled what was described for fifteenth century Kai: a failing polity and greater localization of power. There are no straightforward ways of measuring how self-sufficient and self- governing a medieval village was, but as Hitomi Tonomura has demonstrated, the source of local shrine funding provides a good indicator of the level of village self- sufficiency in this period.74 Under the Estate System, landlords largely funded local shrines. They did so through monetary and material gifts, and by offering significant rebates to villagers for the purpose of shrine festivals, maintenance, and remodeling.

Fifteenth century villages paid for shrines themselves. Tonomura’s study of the villages of the Tokuchin 得珍 district of Ōmi Province indicates that by the 1420s, the shrine association (miyaza 宮座) took up the largest portion of village expenses.75 Funding for shrine upkeep, festivals, rituals, and the priests’ stipends came wholly from local sources. These villages had belonged to an Enryakuji 延暦寺 estate, one of the most prestigious temples in all of Japan, since 1279. Like many elite landlords though,

Enryakuji’s influence waned in the fourteenth century and eventually disappeared. The same kind of local, independent funding of tutelary shrines can been seen in fifteenth century Anji 安治 village of Ōmi, another former Enryakuji holding.76

74 Hitomi Tonomura, Community and Commerce in Late Medieval Japan: The Corporate Villages of Tokuchin-Ho (Stanford: Stanford University Press, 1992), 9. 75 Ibid., 9–10. 76 Sano Chū-kinsei no sonraku to suihen no kankyōshi, 220.

Local processes were instrumental in the formation of a reconstituted sociopolitical order towards the end of the sixteenth century. During the civil war, the “self-help,”

(jiriki kyūsai) phenomenon expanded, essentially amounting to self-governance. This experience through some one hundred and fifty years of conflict had a lasting impact upon society. Even after the establishment of territorial states (domains, ryōgoku 領国) by regional lords, key elements of autonomous, self-help style administrative practices endured locally. There were notable modifications which allowed this kind of local independence to function without widespread disorder. The crucial step in achieving that balance was the emergence of a vested interest in maintaining stability across the sociopolitical spectrum. Basically, warlords left local communities to their own devices.

They intervened to help mediate and resolve disputes that could not be handled locally, thereby preserving social order. In exchange, locals surrendered a portion of their production and accepted daimyo authority as legitimate.

Unlike disputes in the Estate system, warlord regimes featured regular, institutionalized means for dealing with such conflicts. These included, but were not limited to: law codes and courts that were ostensibly applied universally, third-party local officials tasked with investigation of the actual situation, a judicial system in which status mattered less, and was subject to a new discourse of the “public good” (kōgi), and an overwhelming emphasis upon stability (sometimes at the expense of other concerns) which resulted in the establishment of new precendents that were utilized for the next several centuries. The following chapters examine how the above characteristics were achieved in late medieval Kai. By focusing upon a single community, we can trace the

evolution of these contests over the course of the civil war era. I turn now to disputes in late medieval Yoshida Village 吉田郷.

II. Fighting over Nature: Resource Disputes in an Age of Instability

The preceding chapters have argued that localized conflict was the most widespread and influential kind of strife during the civil war era. These disputes were overwhelmingly related to local environmental management. Contests over access to and distribution of local resources were particularly contentious. Resolving such altercations without violence turned out to be essential to the process of ending civil war instability. Moreover, the mechanisms constructed to do so formed the basis of the very stable and durable early modern sociopolitical order. Chapter 1 outlined major changes that increased local autonomy and concentrated a greater percentage of local production outside of traditional power centers in the archipelago. As a result, locals had greater incentive to asset their independence and manage their own affairs.

Conversely, elites’ interest in maintaining order in local areas evaporated. This caused the task of dispute resolution to fall solely into the hands of disputing parties themselves. Gone were the mechanisms for mediation that were held in place by the central-regional patronage networks of the Estate System. In the fragmented, cutthroat atmosphere of war, it is not hard to imagine that direct negotiation with a rival would be quite difficult. Indeed, violence frequently erupted over what were considered to be essential resources. This, more than anything else, was the cause of local disorder, and overall instability that plagued the Japanese islands in the Warring States era. Breaking out of a cycle of repeated clashes over competing claims to local resources proved decisive in reestablishing stability. It was not the conquests of samurai warlords such as

Takeda Shingen, but the evolution of local contests over environment which proved transformative.

Rebuilding a stable system of local environmental management depended upon two seemingly opposed trends: de facto local control and effective central (domain-level) authority. Local control resulted from a marked increase in local autonomy and encompassed a variety of economic and political activities. In terms of environmental management, these activities rendered local residents themselves the effective owners of the places in which they lived and worked. In other words, occupation became the basis for claims to resources. The effectiveness of warlord governance was, from the local perspective, dependent upon the daimyo’s ability to mediate and enforce dispute settlements over these affairs. Local people accepted daimyo authority because it functioned primarily as a guarantor of local claims. Because this role had to be demonstrated in actual practice, it took some time for the relationship between local and regional lord to develop.

From the warlord perspective, local administration was part of a larger geopolitical strategy. Daimyo had to constantly balance military strategy, political alliances, control of retainers, border defense, and others, all while finding ways to tap into local production and minimizing conflict. There was no way Warring States daimyo could impose strict oversight throughout the domain. Instead, they incorporated territory by endorsing local practice. Warlords confirmed local rights in exchange for submission in a handful of key matters. This did subject local people to daimyo rule, however it left local autonomy largely undisturbed. The significant, and ultimately transformative result of this style of territorial administration was that locals gained a powerful patron who could endorse their claims and defend them against competitors.

The Takeda built their domain by functioning as local guarantor and dispute moderator. Their interest in maintaining order and tapping into local productivity actually dovetailed with the desires of local communities to assert their claims. The relationship which developed as a result brought a new kind of stability to Kai. Takeda dealings with subordinates followed a pattern in which confirmation of rights (which here were primarily claims to resources, property, or commercial privileges) was granted in exchange for a specific type of service.1 Tensions between the Takeda and their various subordinates over the terms of submission certainly existed. Tensions between locals themselves were more potentially disruptive. The Takeda could not insert themselves into these local disputes and dictate terms heavy-handedly. Instead, they relied upon local solutions, usually opting for the least intrusive method of resolution and using their coercive power to prevent violence. As they did so, the decisions rendered became the new precedent, remaining influential into the nineteenth century.

Because resource disputes were so fundamental to late medieval society, contemporaries used specific names to describe each according to established types.

Divided into water disputes (suiron 水論), forest disputes (sanron 山論), and field disputes (noron 野論). These words entered the lexicon in the medieval era, and their appearance in historical sources increased dramatically in the late fifteenth and early

1 Such arrangements between warlords and their hereditary military vassals have received the most scholarly attention. Nonetheless, the Takeda established this same basic relationship with all of their subjects, regardless of occupation. Details of the sociopolitical hierarchy established in the Takeda domain are examined in chapter 3.

72 sixteenth centuries – precisely the period of time in which local resource disputes played such a pivotal, transformative role.

The categories employed reflect how medieval people understood the environment as one of three sets of resources: waterways, developed land, and undeveloped land.

Waterways and disputes over them (suiron) were the most basic. The first use of the term suiron in the historical records comes from the mid thirteenth century anthology

Kokon chomonjū 古今著聞集.2 A collection of stories and local customs compiled by

Tachibana no Narisue 橘成季 (unknown), governor of Iga Province 伊賀国, Kokon chomonjū describes a water dispute between local residents of neighboring Estates.

Interestingly, in this case the proprietors of the respective Estates got involved and managed to work out a settlement, indicating that at least in this part of Iga, functional local administration still existed. Water disputes such as the one described appear to have been less common in the Heian and previous eras. However, they increase drastically in frequency from the time of Kokon chomonjū onwards, which is widely considered to be the era when Estate proprietors’ hold on provincial territory began eroding rapidly.3

Water disputes were unsurprisingly the most common type of resource conflict in civil war Japan. Aside from its obvious necessity for sustenance and agriculture, production of all kinds of goods depended upon a substantial supply of water. This included all manufactured goods of the medieval age – everything from rope, to paper,

2 The anthology is thought to have been completed in 1254. 3 Kokushi daijiten henshū iinkai, ed., Kokushi daijiten vol. 14 (Yoshikawa Kōbunkan, 1979), 341. Almost exactly one century later, the 1352 hanzei edict discussed in chapter 1 can be considered a practical acknowledgement of central elites’ loss of functional influence over provincial landholdings.

73 to roof tiles. For instance, in Kōfu, the major urban center of Kai Province, lacquer and cotton cloth became major export goods in the late sixteenth century.4 Both are water intensive to produce. Kōyō gunkan 甲陽軍鑑, an account of Takeda Shingen’s exploits compiled by former retainers in the early seventeenth century, describes cotton agriculture and cloth manufacturing in late medieval Kōfu.5 Shingen himself sent gifts of lacquer ware as part of envoys to Nobunaga in Eiroku 永禄 11 (1568) and Genki 元亀 2

(1571).6 These gifts were almost certainly locally produced, as the Takeda lord would have wished to present specialty items from his jurisdiction.

The second most common type of resource dispute in Kai were forest disputes. The word sanron literally means “mountain argument,” however in this case mountain functioned as a shorthand description for undeveloped land. That term meant any area not under cultivation (whether wet, dry, or slash and burn field) in premodern times.

Undeveloped land was not necessarily forested, but forests were extremely valuable to medieval people and were the most frequently contested kind of undeveloped land in

Kai. For that reason, the translation of sanron as “forest dispute” is appropriate for the purposes here. Nevertheless, it was the undeveloped nature of the land that defined these types of disputes. A sanron could occur over a piece of flat land with no trees on it.

4 Long considered a seventeenth century import, evidence of cotton production has been discovered in late sixteenth century Kai. Note that this is the actual cotton plant (momen or kiwata; written as 木綿 or 木棉), and not the cotton-like silkworm outer cocoon layer referred to by the same name in premodern times (now distinguished as mawata 真綿). Yamanashi kenshi tsūshi-hen 2, 601. See Map 2.1. 5 Sakai Kenji, ed., Kōyō gunkan taisei, 7 vols. (Tokyo: Kyūko Shoin, 1994). 6 Yamanashi kenshi tsūshi-hen 2, 603.

When forest disputes did occur, disputing parties varied. Generally, there are more examples of village versus local landlord sanron in the late medieval period, while village versus neighboring village predominated in early modern times.7 As with water disputes, forest disputes became increasingly common in the late medieval and early modern eras. Competition for resources intensified throughout the medieval period due in no small part to population increase.8 Larger and denser rural communities characteristic of the fifteenth century onwards contributed to more well-defined borders and fewer “unclear” or “unclaimed” spaces. The term sanron itself appears to have entered the lexicon in the early sixteenth century, at a time when this kind of competition was growing more intense, and when institutions to deal with conflict had failed.9 The related terms iriai 入会 “mutual use” and iriyama 入山 “mountain entry” also appeared at this time.10 In other words, the discourse associated with undeveloped land and rights to its use was established in the early sixteenth century, when these issues were being contested by residents all across the archipelago.

People relied on undeveloped land for a huge range of necessities. These areas were essential as a cache of raw materials for local residents. Since Kai has very limited flat land and lowland areas, even by Japanese standards, almost all undeveloped land in the late medieval period was forested. Major resources extracted from forests included fuel, timber, fertilizer, and to a lesser extent fodder for livestock. Horses were the most common draft animals. From at least the eighth century, Kai was known as good horse

7 This reflects a prevailing trend in early modern Japan in which most local landlords had become city dwellers and controlled rural property in name only. 8 Farris, Japan’s Medieval Population, 6–7. 9 Nihon daijiten kankōkai, ed., Nihon kokugo daijiten vol. 19 (Tokyo: Shōgakukan, 1972), 580. 10 Kokushi daijiten vol. 1, 819–20.

75 country, and the Takeda military became famous for cavalry.11 It is difficult to say if the reliance on horses in Kai was greater or smaller when compared to other regions, but it is clear from historical materials that horses were an integral part of the Takeda military and domainal administration. Takeda lords established an extensive communication and transport network of “post horses” (tenma 伝馬) not unlike the

Pony Express.12 Indicating that the animals were owned and used locally is the fact that local areas were required to provide a set number of animals for post horse service every month. Kai did offer several natural advantages for horse-raising. The many rolling mountain foothills made favorable grazing areas. High elevation and generally cooler temperatures were good for the animals as well. Graze land was considered undeveloped and when conflict over use rights occurred they would be considered as forest disputes. While we do not have a direct account of such a dispute from late medieval Kai, there is a Takeda edict of 1544 banning unauthorized grazing on a parcel claimed by Kanezakura Shrine 金桜神社 of Kōfu.13

11 Both Kojiki 古事記 and Nihon Shoki 日本書紀 relate stories depicting the “swift horses” of Kai as early as the fifth century. The legend of Prince Shōtoku 聖徳太子 (574-622) includes an episode where the prince travels to Kai on a mission to find the best horses in all the realm for a gift to Emperor Suiko 推古天皇 (554-628). The prince one day decided to test the locally-born stock and mounted the “black pony of Kai” (Kai no kurokoma 甲斐の黒駒). To his delight, the animal exceeded expectations, reportedly leaping as high as the heavens and traversing across the entire province and back in only three days. These tales are collected in Shōtoku Taishi denryaku 聖徳太子伝暦, a tenth century anthology of the exploits of Prince Shōtoku (574-622). For many years, scholars considered the Heian aristocrat and poet Fujiwara Kanesuke 藤原兼輔 (877-933) to be the compiler. Recent analysis has cast doubt on that assessment, and the origin of the document is unknown. 12 The post-horse system began in Takeda Nobutora’s day. It is considered by many scholars an important indication of Takeda supremacy in the region, solidified in the 1520s and 30s. Yamanashi kenshi tsūshi-hen 2, 440. 13 Yamanashi-ken shihensan iinkai, ed., Yamanashi kenshi shiryō-hen 4: chūsei 1, kennai monjo (Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 1999), 588.

Though brief, the 1544 edict reveals claims to undeveloped land in a state of flux.

The order’s existence shows that people not belonging to Kanezakura Shrine were grazing their animals on this land, and that the temple requested the regional lord,

Takeda Shingen at the time, to ban the practice. Kanezakura needed to keep the land undeveloped as a food source for livestock. It was, generally speaking, a source of raw material. This was the same reason communities of all kinds fought to keep various nearby areas undeveloped. Who had the rights to maintain undeveloped parcels, excluding other kinds of uses, was a contentious issue in the late medieval age. We can see from this example that while shrine personnel felt they had exclusive claim to this parcel for grazing, others did not. Resolutions to problems such as these were worked out across Kai in the mid sixteenth century, the same time that the edict under scrutiny was disseminated.

Undeveloped land was not only a food source for livestock, but people. A very small percentage of the land area in Kai is suitable for cultivation. Of that, the lowlands of the

Kōfu basin are the only lands in which wet fields can be reliably tended under premodern techniques (see Map 2.3). Slash and burn agriculture therefore continued to be an important part of local food production well into the nineteenth century.14 And when all else failed, residents turned to forests as emergency food stores. People regularly ate leaves, roots, and bracken during times of scarcity. In sum, a late medieval community in Kai Province could not survive without a reserve of undeveloped land readily available. Across the archipelago, only a handful of large urban centers

14 Yamanashi kenshi tsūshi-hen 2, 600.

Shinano

Kuninaka Kōfu basin (Takeda)

Kōfu

Gunnai (Oyamada)

Main route Yoshida through region Koma (Anayama)

Sagami

Suruga

Map 2.1: Regions of Kai Province (Yamanashi Prefecture)

which were plugged into vigorous trade networks could do so. Kōfu would not achieve such status until the closing decades of the sixteenth century.

The third kind of premodern resource dispute was that over cultivated land. More precisely, over the borders between different claims to cultivated areas. Because of the emphasis upon agricultural plots by the early Japanese state, conflicts over cultivated land have the longest documented history. Despite that fact, as with water and forest disputes, a new term came into usage in the early sixteenth century to describe these disputes.15 Noron increased in frequency over the late medieval and early modern eras as well. These types of disputes were less common in late medieval Kai, due to the comparative lack of arable land. Areas that were under cultivation would have fallen under the rubric of a field dispute, although the division is somewhat artificial as issues related to all three types of disputes were closely intertwined. Nonetheless, field disputes in Kai were most numerous in the Kōfu basin, over wet-rice cultivation.

Outside of the Kōfu basin, dry fields predominated. Major crops included foxtail millet

(awa 粟), baryard millet (hie 稗), potatoes (imogara 芋柄), taro (akame imo 赤目芋), wheat (komugi 小麦), barley (mugi 麦/ōmugi 大麦), radishes (daikon 大根), soybeans

(daizu 大豆/azuki 小豆), and gourds (uri 瓜). Farmers, especially those attached to temples, used the many rolling hills along the base of mountains to grow tea. The aforementioned commercial crops, cotton and lacquer trees, were a large part of regional agriculture as well.16

Examination of several case studies of well-documented water and forest disputes in late medieval Kai reveals the emergence of what I argue are the two key ingredients for the reestablishment of stability: de facto local control plus effective enforcement of

15 Kokushi daijiten vol. 6, 629. 16 Yamanashi kenshi tsūshi-hen 2, 600.

79 local claims. The examples that follow show an evolution from direct conflict between local disputants that included violence to a regularized judicial process mediated by an elite third party, which used local precedent as the guide for decisions. In the intervening stages, we can see a semi-direct, half-procedural, half-confrontational form of dispute resolution as this evolution takes place. These examples likewise illustrate the shifting logic of claims and rights in the late civil war era. Occupation essentially became ownership by the early seventeenth century. Residents were afforded wide latitude in their management of the local environment. As long as basic conditions were met, such as the forwarding of tax payments and no open rebellion, there was limited oversight from above. That can be contrasted with the period of quite vigorous oversight of local areas by agents of central elites as described in the previous chapter.

The following section highlights an area of Kai for which evidence pertaining to resource disputes is comparatively robust. This region, known today as Fujiyoshida 富

士吉田, was a dynamic crossroads of travelers and trade in the medieval age. It was also home to a thriving community of tradesmen, who relied upon local resources for their livelihoods. Late medieval Yoshida illustrates how the large-scale evolution described above proceeded in a particular place (see Map 2.2). It shows how contests over local environment reconfigured a variety of sociopolitical relationships at multiple scales.

A. People and Place of Late Medieval Yoshida

In analyzing resource conflicts in Yoshida, it is useful to define interactions between groups at three different geopolitical scales. The first and primary subjects of this chapter are local-level residents. This was a varied group socioeconomically. Some were wealthy landowners, others hereditary servants. In this case the term “villager” is employed as a catch-all to encompass those residents of Yoshida who had neither status nor office, distinctions that could be independent of wealth. In Kai, this group was referred to collectively as the “lower people” (jige 地下, jigenin 地下人, jigeshū 地下

衆)17 However wealthy or locally influential they might be, individuals in this group were not elite, nor were they members of the ruling class.

People who possessed status were of three types in medieval Yoshida: religious specialists, heads of artisan guilds, and military retainers. People of status tended to be wealthier, although that was not always the case. They enjoyed special designation within the community and were usually the only ones to possess surnames. Because of their attachment to elite institutions or high-ranking military lords, they were afforded honorifics when addressed in materials written by daimyo officials.18 Religious specialists included monks, nuns, and shrine priests. These were educated people who managed the considerable resources of the numerous shrines and temples that populated Yoshida. Leaders of carpentry, lumberjack, and shipping guilds also resided

17 Note that the terms jige, jigeshū, and jigenin were sometimes used to describe low-level hereditary warriors in other regions. This was not the case in Kai. 18 For instance “lord” (dono 殿, shu 主), “master” (kata 方), or simply “monk” (bō 坊) in the case of low-level temple personnel.

81 in Yoshida. Those employed as military retainers doubled as local landlords.19

Collectively, they was known as “local samurai” (jizamurai 地侍), although this term was not employed in Kai as frequently as in other regions. Medieval Yoshida had three important local samurai lineages: the Kawaguchi 河口, Watanabe 渡辺, and Kobayashi

小林.

Above local status-holders stood county-level elites. These were influential warrior lineages who controlled large areas of a province, usually one or more counties (gun

郡). Called “men of the province” (kokujin 国人), local warriors ostensibly owed allegiance and military service to members of this group. Kai Province was dominated by three such lineages in the late medieval period, each controlling two to three counties: Takeda, Oyamada 小山田, and Anayama 穴山. The Oyamada controlled the southeastern third of Kai, called Gunnai 郡内, where Yoshida is located (see Map 2.1).

To the southwest, the Anayama dominated the Kawauchi 河内 area of the province.

Located in the northern half of Kai, the Takeda held the most productive area known as

Kuninaka 国中.

The Takeda transitioned from county elites to regional warlords in the first few decades of the sixteenth century. Supreme over even rival kokujin lineages, it was at that time that the Takeda became provincial-level elites. This third sociopolitical tier consisted of top warlords who controlled autonomous territorial states. They did not

19 Warriors tended to be landowners in most regions of premodern Japan. One reason is that it required considerable wealth to be able to train and outfit as a military specialist. Many local warrior lineages had parleyed positions as on-site Estate managers (jitō) into hereditary landholdings over several generations.

82 owe allegiance to superiors, but rather commanded the allegiance of both local and county-level warriors (ostensibly, at least). The Takeda lineage had a long history of command in Kai Province, from their position as military governors during the early medieval age. They were able to transition into independent territorial warlords, eventually ruling the province and more in their own right.20 After more or less pacifying Kai in the 1530s, the Takeda’s main rivals became other warlords based in other regions. Because the concerns of this top-level class were on a much larger, regional scale, their interest in localized environmental contests was very different from that of local people. It was warlords’ priority in maintaining internal stability within the domain that stimulated a delegation to local management of resources.

The key resources at stake in late medieval Yoshida were influenced by the area’s major tradespeople: religious specialists, pilgrimage guides, lumberjacks, carpenters, overland shippers, and farmers. All of these groups relied heavily upon water. Basic sustenance aside, water was a necessary ingredient for all kinds of craft production, not to mention the raising of livestock or crops. Transportation figured prominently in most of the trades listed above. Livestock appear to have been relatively prevalent in medieval Yoshida, based upon the number of disputes related to grazing, the frequency of levies on post-horses, and references to packhorses in historical materials.21 Guides, lumberjacks, shippers, and farmers all relied upon draft animals, predominantly horses

20 In his seminal study of premodern political economy, John Hall classified three types of Sengoku daimyo based upon origins. First were those, such as the Takeda, who made the jump from military governor to that of Warring States lord. Second are powerful vassals of military governors who overthrew their lords and assumed regional dominance. Finally are the group which rose from obscurity rule territory independently. See Hall, Government and Local Power in Japan. 21 The post-horse system is discussed in chapter 3.

83 in Kai. Consequently, areas for grazing these animals, what would be considered

“undeveloped” land, were deemed essential. Developed land, too, was a precious, relatively limited commodity in Yoshida, in particular for religious specialists and farmers. Feeding the huge number of temple and shrine personnel required considerable resources. While small time farmers did not practice water intensive rice agriculture, they did extensively cultivate potatoes and a kind of irrigated wheat. Each of the main occupational groups depended upon forest resources, for varied purposes.

Everyone used forests for various raw materials (in rope, for instance), fuel, fertilizer, and as an emergency food source in times of scarcity. Extracting these resources favors a particular type of growth: fast-growing, small to medium woody plants and thick underbrush. Lumberjacks relied upon a different kind of forest for their trade, one that that featured large, old-growth trees that could be converted into construction materials. Many Yoshida farmers practiced slash and burn agriculture in addition to planting established fields. These competing uses could place different demands upon a single landscape.22

22 The “multiple use” system has been analyzed by a number of environmental historians in recent years. It was based upon overlapping or staggered usage rights to particular resources and reached its full expression during the Tokugawa Period. Notable works include Conrad D. Totman, The Green Archipelago: Forestry in Preindustrial Japan (Berkeley: University of California Press, 1989); Sano, Chū-kinsei no sonraku to suihen no kankyōshi; Nishikawa Kōhei, “Sengoku-ki Kai no kuni ni okeru zaimoku no chōsetsu to yamazukuri,” Takeda-shi kenkyū 39 (2009): 1–20; Philip C. Brown, Cultivating Commons: Joint Ownership of Arable Land in Early Modern Japan (Honolulu: University of Hawaiʻi Press, 2011); Bruce L. Batten and Philip C. Brown, eds., Environment and Society in the Japanese Islands: From Prehistory to the Present (Corvallis: Oregon State University Press, 2015).

Lake Kawaguchi

Lake Shoji Katsura River

Yoshida Lake Motosu

Lake Sai

Lake Yamanaka

Mt. Fuji

Map 2.2: Yoshida and Environs

Located in Gunnai, along one of the few corridors through Kai Province, Yoshida was a crossroads in medieval times (see Maps 2.1 and 2.2). The village and modern city are nestled in the northern foothills of Mt. Fuji, near the border between Kai and Sagami 相

模 Provinces. The location is at elevation, with the settlement itself resting at nearly three thousand feet. It features cold winters and hot summers. Rice agriculture is nearly impossible in because of the cold and short growing season. In premodern times, the area was extremely reliant upon snow from Mt. Fuji for its water supply. Mountain runoff flows from numerous streams into five large (by Japanese standards) lakes arranged in a semicircle along Fuji’s northern base. The waters from lakes Motosu 本栖,

Shōji 精進, Sai 西, and Kawaguchi 河口 drain to the northwest, eventually feeding the

Katsura River 桂川, a major tributary of the Sagami River 相模川. On the other side of

85 the village, Lake Yamanaka 山中 drains into the Katsura River too, from the southeast.

Yoshida is therefore located at the mouth of the Katsura River valley, at the confluence of several water flows.

Because of its location, Yoshida functioned as the southeastern entry and exit of Kai

Province. Transportation and commerce were the lifeblood of the medieval Yoshida economy. We can see from Takeda concessions to local tradespeople that travel and trade-related inducements were key. These included road construction and maintenance, low or exempted tolls, and open markets. A religious hub, nearby Mt. Fuji brought in a huge number of tourists annually (as it still does). The pilgrimage

“industry,” for lack of a better term, was a key component of the local medieval economy (and still is). Guides offered escorts up and down the mountain, as well as a plethora of specialized products and services associated with the deities of the mountain. Many Yoshida residents were attached in some way the several Nichiren 日

蓮 temples or Sengen Shrines 浅間神社 of the area. 23 Institutions such as these were the financial centers of the medieval world. They sponsored markets, in turn providing other tradespeople a venue to sell their goods and services. As hosts, temple administrators were keenly interested in commodity prices. Local monks in Yoshida diligently recorded fluctuations in the value of one crop or another, as well as the

23 Osano, “Takeda ryōgoku no dogōsō to chiiki shakai," 331. The tutelary shrines (chinju 鎮守) are Kitaguchi hongū Fuji sengen jinja 北口本宮富士浅間神社 and Omuro sengen jinja 御室浅間神社 (see Map 2.3). Sengen Shrines (called Asama shrines further north) venerate Konohana no sakuyabine 木花咲耶姫, the principle deity of Mt. Fuji. Unlike previous eras, the temples and shrines in Yoshida were controlled and managed locally. They were not satellites of the elite temples of the capital as had been common under the estate system. Little coordination existed even between institutions of the same sect. They were, like the settlement itself, largely local and independent. Lists of donors to the tutelary shrines of Yoshida indicate a parochial funding base.

86 numbers of pilgrims and anything that disrupted trade. The accounts left behind reveal a place that was heavily dependent upon imports. Local production had to be supplemented, meaning that disruptions to either local agriculture or trade could have a devastating impact upon the food supply. Indeed, Yoshida appears to have been quite vulnerable to famine in the late medieval age. Nonetheless, it was a site of movement and exchange. It was what Amino Yoshihiko considered an “urban space” where people and products from various regions intermingled. 24

Yoshida was a geopolitical crossroads in the civil war era as well. A border area in during the civil war, it sat on the edge of the Takeda domain near the territory of the

Go-Hōjō 後北条 of Sagami. It was subject to several invasions by neighboring armies during the first half of the sixteenth century. Merchants, artisans, and even local samurai maintained relationships with Hōjō warlords despite technically being within

Takeda territory proper from the 1520s on. Both the Oyamada and Kobayashi lineages alternated periods of resistance to and cooperation with the Takeda before ultimately submitting. But that process took some time. The Kobayashi were not integrated as

Oyamada vassals until around 1530, and it would be another decade until Oyamada themselves were firmly under Takeda suzerainty.25 The fluid political situation meant a great deal of uncertainty for local residents. Essentially, they could not rely upon external forces to endorse and protect their claims until there was a demonstrated

24 Amino contrasts premodern urban spaces with rural settlements which were largely insulated and self-sufficient. His major point is that although there were only a handful of cities in premodern Japan, there were many “urban spaces.” Amino, Rethinking Japanese History , 74. 25 Shibatsuji Shunroku et al., eds., Takeda-shi kashindan jinmei jiten, 2015, 315–16; Yamamoto Takeshi and Owada Tetsuo, eds., Sengoku daimyō kashindan jiten: tōgoku-hen (Tokyo: Shin Jinbutsu Ōraisha, 1981), 382.

87 willingness and ability to do so. There was no reliable way of knowing who could effectively offer what Mikael Adophson and J. Mark Ramseyer have called “basic government services,” or what the nature of those services would be in this unstable atmosphere.26

Almost all of what we know of late medieval Yoshida and its people comes from the writings left behind by local monks. Written and compiled under unclear circumstances by Nichiren priests, the temple chronicle Myōhōji-ki provides extremely valuable information on this very sparsely documented era.27 Moreover, the document’s perspective offers insight into the affairs most pertinent to the local Yoshida community. The tumult of interregional war reverberates through these accounts, but references to campaigns, battles, and the goings on at Takeda headquarters in Kōfu are distant.

When the wars do directly enter local space, armies on any side represent serious threats. For instance, in the midst of a huge mobilization of forces for a 1540 invasion of

Shinano, Yoshida monks wrote of “warriors all over the place causing problems.”28

Reigning in the destructive, potentially harmful actions of their own followers was a

26 Mikael Adolphson and J. Mark Ramseyer, “The Competitive Enforement of Property Rights in Medieval Japan: The Role of Temples and Monasteries,” Journal of Economic Behavior and Organization 71 (2009): 660–68. Adolphson and Ramseyer highlight property claims and dispute adjudication as critical components of basic government services. 27 The chronicle was likely begun by Nikkoku 日国 (d. 1528), a fairly well known local priest. It was later expanded by subsequent authors to cover the years 1466-1561. While authorship is unclear, it is clearly from the vantage point of Yoshida. It may have been written at nearby temples Myōhōji or Jōzaiji 常在寺. An alternate version called Katsuyama-ki 勝山記 is extremely similar, and likely produced from the same source materials. Debates continue over the origin of this document and the relationship between the two versions of the chronicle. I have consulted both in the research presented here. 28 Myōhōjiki, 32.

88 major challenge for the Takeda and other regional warlords. In 1561, Shingen issued a sealed order to the village leaders in Yoshida confirming that Takeda troops (gunzei 軍

勢) were barred from cutting wood in nearby Suwa 諏方 forest.29 For much of the civil war era, samurai armies practiced a kind of “provision as you go” method on invaded territory and within the domain. These forces relied heavily upon arson, kidnapping, and scorched-earth tactics. Because of their widespread use, historian Wayne Farris considers these tools to have been the most effective weapons in medieval warriors’ arsenals.30 Takeda Katsuyori wrote a letter apprising retainers of the status of a 1574 campaign against Tokugawa Ieyasu in Tōtōmi (modern Shizuoka) and noted: “homes of the people have been burned, with not one remaining, and the rice plants, too, have all been cut up and discarded.”31 We know from local sources such as Myōhōji-ki that residents sometimes faced such treatment. Conversely, the document shows that local communities were equipped and ready to defend themselves from such incursions.

Yoshida residents did so on a number of occasions, most notably in 1516.32 Facing an invading force from Imagawa of Suruga, villagers retreated to a stronghold on

Unoshima 鵜ノ島, a small island in the middle of Lake Kawaguchi, before successfully launching a counter attack which repelled the aggressors.33 Several locals died in the fighting. Noteworthy as these military encounters are, they are not the main focus of

Myōhōji-ki’s year by year account. Reports on commodity prices, the availability of

29 Yamanashi kenshi shiryō-hen 4, 1055. Not to be confused with Suwa 諏訪 of Shinano Province. 30 Farris, Japan’s Medieval Population, 197, 200, 207. 31 Yamanashi kenshi shiryō-hen 4, 1077. 32 Myōhōji-ki, 16. 33 Fujiyoshida-shi shihensan iinkai, ed., Fujiyoshida-shi shi shiryō-hen dai nikan (kodai, chūsei) Fujiyoshida-shi: Fujiyoshida-shi, 1992), 218.

90 goods, the number of travelers in and out of the area, weather, and crop conditions are the most frequent and detailed topics. When conflict and violence appear, it is usually within the local community itself. Each of these disputes are at their core over local resources. Let us now examine a series of such disputes in detail.

B. Case 1: Reactions to Local Infrastructure Building (Yoshida vs. Kawaguchi,

1475; Yoshida vs. Kobayashi, 1504)

Key sites of civil war conflict for most islanders occurred within local communities themselves. Although collectively the population of late medieval Yoshida was lower class, non-elite people, stratification within that group did exist. The majority of residents owned little or no property, but there were a handful of wealthy lineages who controlled significant landholdings. These included the aforementioned Kawaguchi and

Kobayashi houses, who also happened to be warriors.34 Yoshida was therefore a diverse community of various occupations and socioeconomic levels who nonetheless had to find ways to share local resources. There were customary mechanisms in place to do so, including a highly developed system of managing communal property. These practices were not always agreed upon by all, and wealthy residents in particular might have different preferences for how resources should be distributed. The resulting socioeconomic tension erupted frequently in the Warring States era, in particular during times of crisis.

34 Kawaguchi and Kobayashi were not the only landlords in late medieval Yoshida, but were certainly two of the most significant.

The unpredictable water supply in Yoshida was one persistent cause of trouble.

Droughts and floods repeatedly led to conflicts. Water disputes frequently initiated modifications to irrigation infrastructure, an extremely contentious undertaking. The clearest examples of such instances in late medieval Yoshida indicate that the community tended to fracture along socioeconomic lines, pitting the wealthiest, landowning few against the rest of the community. But this was not an era when social superiors could easily assert themselves. Anything contrary to the collective will faced coercive force. In 1475, the residents of Yoshida killed a local landlord and eliminated his lineage, thus removing a disruptive competitor for local resources. Consensus prevailed only around infrastructure projects whose benefits could be distributed more broadly. Wealthy residents could use these projects to increase their standing within the community while maintaining harmony, as we will see from a 1504 example.

The year Bunmei 7 文明七 (1475) began with heavy rains and spring flooding in the

Yoshida area. The terse entry from Myōhōji-ki states it was “not a prosperous year.”35

Then, in the twelfth month, an armed force from Yoshida rose up and killed one lord

Kawaguchi 川口殿.36 All of this occurred on the heels of a widespread famine that had begun just two years prior 1473.37 Taken together, it is safe to say that the mid 1470s were not a prosperous time for Yoshida. By piecing together several historical accounts,

35 Fuku oyobazu mōshi [sōrō] 福貴不及申『候』. Myōhōji-ki, 4. 36 Ibid. The force was almost certainly the village militia (wakashū), which is not consistently named in local sources. Later references to the “group of twenty” nijūnin 二十人, appear to be to members of militia leadership. 37 Ibid.

92 it likely that the Kawaguchi were killed over a water use dispute. We can say for sure that the local Yoshida community had the power to kill a local samurai.

Kawaguchi was a local landlord in Yoshida who controlled a very desirable piece of real estate. By virtue of the epithet “lord” (~dono), he was clearly a person of status. Kai

Kokushi 甲斐国志, an early modern historical survey of Kai Province, records a medieval estate (kyokan 居館) on the north shore of Lake Kawaguchi belonging to

Kawaguchi Saemon 河口左衛門 (see Map 2.3).38 This estate sat just north of Zen’ōji 善

応寺, where a horse-riding ground (baba 馬場) still bears the Kawaguchi name

(Kawaguchi yashiki no baba 河口屋敷の馬場).39 That would place it directly next to the local Sengen Shrine, known colloquially as Kawaguchi Sengen 河口浅間. In other words, we can be reasonably sure that this was the location of lord Kawaguchi’s home mentioned in Myōhōji-ki.

This spot had several topographic advantages. Located in a small valley on the shores of Lake Kawaguchi, it offered ready access to water running off the mountains and into the lake. Nestled between the peaks, it was easily defensible, yet readily accessed from the larger settlement at Yoshida. And, it was nearby the major thoroughfares of the region. Kawaguchi had first access to any water draining from the northwestern watershed before it reached the lake. This meant that any sluices, dikes, or catchment ponds at the Kawaguchi estate affected lake supply. Yoshida, further downstream, relied on several water sources, one of which was spillage out of Lake

38 Matsudaira Sadamasa, ed., Kai kokushi (Tokyo: Yūzankaku, 1968), v. 53 no. 513. 39 Fujiyoshida-shi shiryō-hen dai nikan, 160.

Kawaguchi to the south and into several streams running through the village. Lake levels had to be high in order to maintain that flow.

While the provocation is unrecorded, Kawaguchi must have done something that angered his fellow Yoshida residents. Whatever he did, it was serious enough to result in armed conflict and ultimately Kawaguchi’s death. When we consider the context of two years’ worth of famine and destructive flooding, it is easy to see why tensions were so high at this time in Yoshida. Coming as it did directly following a flood, Kawaguchi’s murder could have been part of an ongoing water dispute. Examination of later materials from the area hint at this possibility.

In Bunki 4 文亀四 (1504), a local warrior called Kobayashi Owari Nyūdō 小林尾張入

道 (d. 1535) made some improvements to the irrigation system of his residence.40

Sluices that had been installed by the previous lord were removed, and new ones built at Funatsu 船津 (see Map 2.5).41 From an outlet called tsutsu no guchi ツツノ口, the new channels pulled water directly from Lake Kawaguchi to the Kobayashi estate. This project, like the attack on Kawaguchi, appears to have been a reaction to local environmental crisis. A multi-year drought resulting from dry winters and below average snow melt caused a famine in Yoshida that stretched from 1502 to 1506. 42

40 Also known as Dōkō 道光, this Kobayashi was the second of three to hold the title Owari no kami 尾張守. His father Kobayashi Shōki 小林正喜 (unknown) aka Owari Nyūdō is considered the patriarch of the Funatsu-based, Owari no kami branch of the Kobayashi. Shōki was the first to appear with the name Owari Nyūdō, strongly suggesting that he held the title Owari no kami. This title was likely passed to his son Dōkō, who also took the tonsure and the priestly name Nyūdō and initiated the construction project described here. Dōkō is thought to be the father of a third Kobayashi Owari no kami named Shosuke 少輔 or Miyauchi no suke 宮内助 (d. 1584?). He clashed with Yoshida villagers over resources in the late 1550s several times. See Shibatsuji et al., Takeda-shi kashindan, 316–20. 41 Myōhōji-ki, 10. 42 Ibid., 10–11.

Could Kobayashi’s new sluices have been part of an effort to revamp aspects of the local irrigation system, perhaps to mitigate the ongoing drought and starvation? A frustrating lack of evidence prevents any firm conclusions, but this reconstruction of events does fit an observable pattern in Yoshida’s subsequent water disputes: disruption to water supply – hardship (indirectly or directly as a result) – construction/modification to local infrastructure – and then usually, conflict. That there was no conflict in 1504 indicates Kobayashi enjoyed broad support in Yoshida, or that other locals lacked the wherewithal to stand up to him.43 The latter is unlikely considering that Kobayashi influence was in its infancy at the time. Yoshida residents did in fact challenge

Kobayashi, vigorously at that, in later decades when the Kobayashi were even more powerful than they were in the early decades of the sixteenth century.

The Kobayashi were influential, wealthy local residents in Yoshida since at least the

1470s. But in the early sixteenth century, when this construction project took place, the

Kobayashi were just emerging as a lineage of hereditary warriors. They participated in many campaigns eventually becoming incorporated as permanent retainers to the

Oyamada, and subsequently Takeda, by midcentury.44 Their first appearance in the local historical record in 1475 coincides with Kawaguchi’s disappearance from it.

Kawaguchi may have been the “previous lord” mentioned in the 1504 entry of Myōhōji- ki.45 It is not far-fetched to consider that the Kobayashi took advantage of Kawaguchi’s

43 Otherwise he might have been dealt with in the same manner as Kawaguchi. 44 Shibatsuji et al., Takeda-shi kashindan, 315. 45 While the Kawaguchi residence was on the opposite side of Lake Kawaguchi from Funatsu, where Kobayashi removed sluices of the “previous lord,” this does not preclude Kawaguchi involvement in said sluices. Local documents do not mention any other landlord in this area during the intervening years between 1475 and 1504. It is highly unlikely such a person would be overlooked in a local chronicle such as Myōhōji-ki. It is a reasonable speculation that Kawaguchi was the “previous lord,” though there are many other possible interpretations.

95 elimination to increase their standing in Yoshida.46 In an interesting contrast, the same year villagers eliminated Kawaguchi, Kobayashi Nyūdō made a significant contribution

(kishin 寄進) of property to Fuji Omuro Sengen Shrine on the third day of the eighth month.47 As the tutelary shrine, a gift to this institution had a certain civic function – at least in the sense that shrine resources served the community as a whole. While we do not know what Kawaguchi did to draw his neighbors’ ire, it must have been deeply unpopular. Kobayashi, on the other hand, ultimately went on to become influential local leaders. Donations to the communal shrine and leadership in sluice reconstruction were surely components of that.

The opposite reactions these local landlords experienced indicate a collective, autonomous local sociopolitical body. Kobayashi must have had something to offer fellow local residents. Historians Sasamoto Shōji 笹本正治 and Osano Asako 小佐野浅

子 have concluded that the 1504 project presented something of communal value, and was thus embraced by the majority in Yoshida.48 Based upon local archaeological sites, it was likely a project to build irrigated fields for the cultivation of “wet wheat”

(mizujake-mugi 水掛麦) for Omuro Sengen in 1504.49 These kinds of projects were the primary methods by which Kobayashi increased their local influence, and ultimately

46 Based upon the favorable local reaction to Kobayashi’s 1504 project, it could even be that Kobayashi was instrumental in leading fellow Yoshida locals against the Kawaguchi in 1475. Limited evidence again forces mere speculation. 47 Yamanashi kenshi shiryō-hen 4, 1105; Shibatsuji et al., Takeda-shi kashindan, 318. The donation was of arable land to the shrine (shinden 神田). See note 40 for information on Kobayashi Nyūdō aka Shōki. 48 Fujiyoshida-shi shiryō-hen dai nikan, 202; Osano, “Takeda ryōgoku no dogōsō,” 329. 49 Wet wheat is a hardy strain that can be grown at the high elevation of Yoshida. It requires irrigation, but not a submerged, wet field as in the case of rice. Yamanashi kenshi tsūshi-hen 2, 599.

96 wealth.50 By providing financing, materials, and manpower for the expansion of arable and other “development” (kaihatsu 開発), the local upper class (dogōsō 土豪層) supported local initiatives. They then gained an influential stake in the management of that infrastructure as a result of their role. An early modern lawsuit from Yoshida indicates that this happened for Kobayashi as a result of the 1504 project.51 In it, villagers assert that Kobayashi Nyūdō (Dōkō), a person of Yoshida, had fields built at

Shimo no mizu 下ノ水 and collected undergrowth from a nearby hillside. Behind this assertion is the implication that because Kobayashi was from Yoshida, the place Shimo no mizu belongs to the village, but it also shows that Kobayashi himself enjoyed special privileges for his instrumental role in improving local infrastructure.

Unfortunately, surviving documentation for these two obscure events leave many details unknown, but they demonstrate a number of crucial points. First, there were serious consequences for going against the collective will in Yoshida. Villagers were willing and able to use violence to resolve disputes, and in fact seem to have relied upon force generally to achieve their goals. Whether or not the 1475 killing of Kawaguchi was the result of a water dispute or not, it shows that civil war era communities had force at their disposal, they deployed that force in local disputes, and they had the capability to eliminate a local warrior landlord.52 Second, local residents acted autonomously in the both the conflict of 1475, and the harmonious construction in 1504. The people involved in these actions were also the ones making decisions about them. No

50 Osano, “Takeda ryōgoku no dogōsō,” 336. 51 Fujiyoshida-shi shihensan iinkai, ed., Fujiyoshida-shi shi shiryō-hen dai sankan (kinsei ichi) (Fujiyoshida-shi: Fujiyoshida-shi, 1994), 556–57. This document is discussed in greater detail below. 52 In fact, an entire lineage.

97 intermediary or higher authority intervened to dictate terms or issue orders. While this may seem like a very basic observation, it is an extremely significant characteristic of the late medieval age. It contrasts sharply with systems of local administration in previous centuries. The legacy of localized administration and accompanying autonomy persisted throughout the early modern era, although it became subjected to specific kinds of restraints. The following examples show how those restraints formed, providing the basis for a territorial state effective at maintaining order.

C. Case 2: Intervention in Infancy (Yoshida vs. Watanabe, 1533)

Let us now consider a clear-cut water dispute in Yoshida which occurred roughly two generations later. Like the 1475 example, it resulted in direct action and violence.

Although no one was killed, that easily could have happened as events unfolded. A 1533 standoff between Yoshida and a wealthy local named Watanabe, likely a warrior, reveals the first indications of an emerging process in dispute resolution. The settlement itself also shows preferential treatment to the local community from elites at higher sociopolitical levels.

Tenbun 天文 2 (1533) was a rough year for residents of Yoshida village. Although fair weather prevailed around the New Year, late spring turned out to be exceptionally dry. Water levels at Lake Kawaguchi ran low.53 In the midst of this drought, a fire broke out on the sixteenth of the third month (April 10 by the Gregorian calendar) and

53 Myōhōji-ki, 27.

98 consumed a large portion of Yoshida. According to monk chronicler, “everything but

Jōgyōji 上行寺 was burned.” 54 To this pious observer, the fact that only a temple remained in town indicated the Buddha’s power. While these statements are most likely hyperbole, we can conclude that this fire resulted in considerable damage. As residents fought through the dry weather and salvaged what they could from burned-out structures, the rain returned. From the fifth to the eighth months, Yoshida was inundated, drought quickly turning into flood conditions. The deluge drowned crops, causing a poor harvest and further hardship.

Under these tough circumstances, “the people” of lower Yoshida (shimo Yoshida no katagata 下吉田方々) got involved in a water dispute with one Watanabe Shōzaemon

渡辺庄左衛門 (unknown).55 Watanabe, like Kawaguchi and Kobayashi, was a local landowner and military specialist. Branches of the Watanabe lineage lived in various locales around Yatsushiro 八代郡 and Koma 巨摩郡 Counties in Kai who served the

Anayama. Although the Shōzaemon mentioned here only appears in this record, a

Watanabe lineage of the Lake Kawaguchi and Yoshida area can be confirmed. A Takeda- issued document of 1567 to a lord Watanabe granted rights to hunt waterfowl along the shores of Lake Kawaguchi.56 These privileges were confirmed to the same lineage in the early seventeenth century by the Tokugawa. Shingen and his successor also issued property confirmation (ando 安堵) and enfeoffment (ateokonau 充行) orders to

54 Kami-Yoshida mina yakuru, Jyōgyōji wa nokosu mōshi 上吉田皆焼ル上行寺ハ残申. The temple no longer exists, but was located within Yoshida. Ibid. 55 Ibid., 28. 56 Shibatsuji et al., Takeda-shi kashindan, 716.

Watanabe in the late 1560s and 1570s.57 The Watanabe estate itself was likely somewhere along the southeastern shore of Lake Kawaguchi.58 We know that

Shōzaemon was a retainer of Oyamada Nobuari 小山田信有 (1488-1541) at this time, because when events of the dispute spiraled out of control, Shōzaemon appealed to the

Oyamada for assistance.59 The Watanabe had gained powerful allies outside of Yoshida and they endeavored to bring that external influence to bear in their local troubles.

The immediate cause of the Yoshida versus Watanabe water dispute is not recorded.

Regardless, it is not hard to imagine how the conditions of 1533, a serious drought followed by flooding, could lead to disagreements over the water supply. The area had been plagued by additional challenges in the few years leading up to the dispute. There was famine, a large number of refugees, and commercial shortages.60 Then came the severe shock to the water supply in the late summer of 1533. In that context, Watanabe made modifications to irrigation infrastructure at his estate. This could have involved construction of new sluices or reconfiguration of existing ones. The nature of these changes is unrecorded, but clearly they had an impact upon the local water system.

Yoshida villagers considered that impact unacceptable and moved swiftly to undo the changes Watanabe had created by destroying the sluices.61 We can conclude with near certainty that is was the Yoshida militia who did the actual digging up and knocking down of these structures.

57 Ibid., 716–18. 58 Fujiyoshida-shi shiryō-hen dai nikan, 250. 59 Also known as Etchū no kami 越中守, Nobuari was the first of three Oyamada patriarchs of that name. 60 Myōhōji-ki, 25–26. 61 Ibid., 28.

100

Unlike the previous case of 1475, county-level elites were brought in to help resolve this matter. Interestingly, when Oyamada Nobuari intervened, he ultimately decided in favor of Yoshida village over his own retainer. After hearing testimony from “many people” (iroiro samazama 色々様々) the Oyamada family head backed Yoshida’s claims.62 Watanabe therefore, in the determination of his superior, had no right to implement this construction project. The village was justified in their forced removal of the structures. In other words, local precedent was respected over change – even a change that was desired by a retainer (Watanabe) of the presiding lord (Oyamada).

This case and its resolution are significant because it entailed both direct action and procedure, and because the local community ultimately won endorsement of their claims by a superior operating at a different geopolitical scale. Although none were killed, this dispute still resulted in violence. Yoshida sent out the militia to dig up

Watanabe’s sluices. This was the initial response – an appeal to potential mediators only came later, and it appears to have been initiated by Watanabe.63 In the beginning this dispute was contested between the two parties directly, and again villagers used force to challenge a local landowning military specialist. How the Oyamada came to intervene in a local Yoshida affair has nothing to do with Oyamada conquest of the place. It was actually their own subjugation to the regional warlord, Takeda, which placed them in such a position. The same year as this dispute (1533), Nobuari built a seventy tsubo 坪 (about 280 square yards) estate in Kōfu, indicating that his family had

62 Ibid. 63 Fujiyoshida-shi shiryō-hen dai nikan, 249–50.

101 become an entrenched Takeda vassal.64 The Oyamada were clearly an integral part of

Takeda regional dominance by this point. And conversely, Nobuari’s political and economic superiority in Gunnai were backed by Takeda power. There was now a vested interest and obligation to keep that area smoothly pacified. In what would anticipate a future pattern, for Nobuari the most convenient way to maintain order in Yoshida was to endorse the claims of the community over those of a local follower. This was crucial for the development of dispute resolution mechanisms. Yoshida witnessed that the county-level lord was willing and able to back their claims. Oyamada could help reign in local landowners such as Watanabe. Yoshida villagers now had a demonstrated reason to seek assistance from county elites in future disputes with local rivals.

D. Case 3: Disorder and Procedure (Yoshida vs. Kobayashi, 1556-57)

The basic pattern established above persisted in resource dispute resolution for several decades. Two new developments emerged in the late 1550s which had a lasting impact. The first was the creation of a real procedure for submitting formal appeals, what amounted to lawsuits, for mediation. Second, locals began making these appeals directly to the regional authority. That helped to solidify a convergence of interest between the very elite (Takeda) and very local (Yoshida), while at the same time curtailing the influence of local- and county-level elites, in this case the Kobayashi and

Watanabe lineages. Furthermore, direct appeals to the regional authority resulted in

64 Myōhōji-ki, 27.

102 continuation of the trend towards giving preference to the claims advanced by the community.

Yoshida village and members of the Kobayashi lineage became entangled in a series of linked disputes over a ten year period involving both water and forest rights. As described above, the Kobayashi arrived in Yoshida in the late fifteenth century and had amassed some degree of wealth and influence by the early sixteenth. When they came into conflict with their neighbors around midcentury, the decisions rendered initiated a transition towards de facto ownership for residents. Rather than claims to resources being linked to office, status, or investiture from a superior, occupancy came to determine control.

By the mid sixteenth century, the Kobayashi lineage had split into two distinct lines: one based on the southern shore of Lake Kawaguchi and headed by Kobayashi Owari no kami (d. 1584?), and another located at Matsuyama 松山 under Kobayashi Izumi no kami 小林和泉守 (d. 1580) (see Map 2.5).65 Both Kobayashi lineages controlled strategically advantageous points in Yoshida. The Owari lineage was based at Funatsu, at an estate right on the southeastern shore of Lake Kawaguchi.66 This placed them upstream of the village for any water flowing out of the lake. Osano speculates that

Kobayashi achieved this position by investing in the development of the local water

65 This Owari no kami was the third and final Kobayashi to hold that title. Izumi no kami, aka Fusazane 房実, Keibuzaemon 刑部左衛門 inherited the title from his father, who died in 1536. The Izumi no kami lineage was originally part of the Owari Kobayashi from Funatsu. When the elder Izumi no kami died, his son built a new estate at Matsuyama. The two branches of the Kobayashi family then began to diverge, and they often came into conflict with one another over the next few generations. See Shibatsuji et al., Takeda-shi kashindan, 320–21. 66 They may have displaced the Watanabe from this prime location. The site now hosts a five star resort hotel.

103 infrastructure, including but not limited to Dōkō’s 1504 construction of wheat fields.67

Their location afforded the Owari lineage a high degree of influence over Yoshida’s water supply. The Izumi lineage controlled the Funatsu checkpoint (sekisho 関所).

Located along an old Kamakura-era (1185-1333) highway, the checkpoint marked the border between Funatsu and Matsuyama, areas that basically correspond to Upper and

Lower Yoshida respectively.68 It stood along the only access route in and out of the area.

Control of this point therefore meant control of traffic through Yoshida. The Izumi

Kobayashi used the checkpoint as a major source of income, charging travelers and shippers a fee to pass through. Pilgrims were a particularly important source of such funding. Kobayashi themselves also engaged in shipping. Both Takeda and Go-Hōjō lords issued shipping licenses (tenma tegata 伝馬手形) to the Izumi Kobayashi in the

1550s.69

The first significant clash between Yoshida and Kobayashi was over water, initiated by construction of a new Kobayashi residence. In 1556, Kobayashi Owari decided to build a new estate and dug wells and built fields for the new complex.70 He had expressed his plans to do so to his direct lord, Oyamada Nobuari (1540-1565), but ran into stiff resistance from Yoshida villagers.71 Because of Owari no kami’s “repeated

67 Osano, “Takeda ryōgoku no dogōsō,” 329. 68 Ibid. Although the Izumi lineage controlled the Funatsu checkpoint, their estate was located at Matsuyama (see Map 2.3). 69 Shibatsuji et al., Takeda-shi kashindan, 320. The fact that the Kobayashi were allowed by to travel between the domains of rival warlords illustrates that in many cases, access to trade overrode daimyo strategic military concerns, a point analyzed in detail in the following chapter. 70 Myōhōji-ki, 44. 71 This Nobuari, aka Yasaburō 弥三郎, was the son of Nobuari Dewa no kami 出羽守 (d. 1552) who also went by the name Yasaburō. He was the third of three Oyamada family heads with the name Nobuari (the other two were Etchū no kami and Dewa no kami). Adding further confusion is the fact that the third Nobuari’s brother Nobushige 信茂 (unknown) was mistakenly called Yasaburō in Kōyō gunkan. They were in fact two different people. See Shibatsuji et al., Takeda-shi kashindan, 231.

104 mistreatment of Yoshida residents (Yoshida-shū ni hibun ōku sōrō aida 吉田衆ニ非分多

ク候間), when construction began, the Yoshida “group of twenty” (nijūnin 二十人) went out and confronted Kobayashi’s followers.72 The encounter does not appear to have turned violent, but ended in standoff. Construction ceased. The Yoshida group and

Kobayashi servants then traveled to the Oyamada estate at Yamura 谷村.73 It was here that the two sides expressed their cases to the young Nobuari, a youth of only sixteen years at the time. It must have been a difficult position for the inexperienced leader.

Kobayashi Owari was several decades his senior, not to mention one of his most influential retainers. After several days, Nobuari failed to render a decision.74

Kobayashi’s envoys returned home, but the Yoshida “group of twenty” pressed on to

Kōfu some thirty miles away. There they presented their case directly to Takeda

Shingen and justified their claims based upon logic or “reasonableness” (dōri 道理).75

Shingen determined that Kobayashi Owari no kami should not build a new estate. He must abandon the project and dismantle whatever had already been completed.76

Yoshida villagers were undoubtedly pleased with Shingen’s decision, and it emboldened them to take further steps to curtail Kobayashi’s influence locally. That same year

(1556) they severed the relationship with Owari no kami in which he served as a community patron. This hierarchical arrangement had a distinctly martial connotation

72 Myōhōji-ki, 44; The "group of twenty" was most likely a cohort of village militia leadership and younger members of the village council who served as messengers and envoys for the council itself. See Fujiki, Sengoku no sahō, 18, 24, 33; Nishikawa, Chūsei kōki no kaihatsu, 102–3. 73 In modern Ōtsuki City 大月市, about twenty miles northeast of Yoshida over a mountain pass. The Oyamada main estate was located at Yamura, and they held a castle at Iwadono 岩殿. 74 Myōhōji-ki, 44. 75 For an analysis of the discourse of reasonableness and its use in disputes, see Eason, “The Culture of Disputes.” 76 Myōhōji-ki, 44–45.

105 in medieval Japan, and was likely related to Kobayashi’s ability to lead local military forces by virtue of his occupation. In such an arrangement, the superior was called yorioya 寄親 (“parent”) and the subordinate yoriko 寄子 (“child”). 77 These terms were used to describe master-retainer relationships within the warrior class as well. We know from other examples that low-ranking warriors were sometimes placed in charge of organizing defense forces in local areas. A local warrior such as Kobayashi would have been in charge of mobilizing the non-professional fighters in Yoshida (in other words, the militia) in times of emergencies. This in fact happened once in 1516 during an invasion by Imagawa forces from Suruga.78 At that time, Dōkō, father of the

Kobayashi Owari no kami in the 1556 dispute, led local military forces to victory.

Referred to as jige, the participants were non-professional ordinary residents of the place. As evidence by their victory, they did have some military proficiency.

Other examples from the Takeda realm show the recruitment of jigenin for participation in campaigns. Takeda offered special one-time property tax exemptions in exchange for service on specific campaigns. On one occasion, an undated document with Shingen’s seal authorized a property tax exemption to the entire village of

Katsuyama 勝山, Yoshida’s western neighbor.79 Property tax exemption is significant because it was the basic perquisite Takeda gave to retainers for military service. Since this exemption was one time, it is highly likely it was given as a reward for battle participation. It makes sense that militias would have an incentive to participate in

77 Kokushi daijiten v. 14, 471–72. 78 Myōhōji-ki, 16; Osano, “Takeda ryōgoku no dogōsō,” 332. 79 Yamanashi kenshi shiryō-hen 4, 1105.

106 battles from time to time when they impinged upon the local area. And warlords could bolster their forces by hiring an on-site group with military experience.

This was completely separate from the lord-retainer relationship with hereditary vassals.80 There is some debate about the exact status and nature of jigenin in late medieval Japan. Some scholars consider them to be local warriors who eventually became members of daimyo vassal bands. The terms jigenin and jizamurai are for this group interchangeable. While that may be true in other regions, it is clearly not the case in Kai. The precise term jizamurai does not appear in late medieval Kai documents, but jigenin and samurai are clearly considered distinct groups. The former were not full time, hereditary warriors. Jigenin lacked status, surnames, and titles, and unlike vassals, conditions of servitude were not regularized. The relationship was not to be in perpetuity. In various Kai documents, jige is used in one of two ways: to describe part- time, local fighters such as the village militia, or as a broader term referring to the ordinary people of a place.

Kobayashi’s position as yorioya to Yoshida was a kind deputizing in which a military professional took a leading role in local defense. However, when Yoshida villagers walked away from the arrangement in the wake of the 1556 dispute, it was an explicit rejection of Kobayashi’s place as an individual of status in the community. It is important to note here that Kobayashi’s standing in late medieval Yoshida was quite different than that of Estate managers (jitō) of previous eras.81 He was recognized as a prominent local citizen and person with specialized skills, not the agent of a remote,

80 Fujiyoshida-shi dai shiryō-hen nikan, 102; Sasamoto, Sengoku daimyō, 261–62. 81 See chapter 1 for a summary of the role of jitō.

107 elite estate landlord. Despite the enmity between Kobayashi and his neighbors, he was still part of Yoshida. Residents made this explicit in the 1635 early modern lawsuit referenced above when they called Kobayashi Dōkō, “a person of Yoshida.”82 The Owari no kami here, likely Dōkō’s son, was a prominent local resident, not a dictator. He and other wealthy community members would have been expected to contribute to the collective just like anyone else.83 And as the ranking retainer in the area, Kobayashi would have been turned to as a conduit to higher ups such as the Oyamada and Takeda, with whom he maintained a relationship. That of course did not happen in this case, indicating that villagers were willing and able to circumvent Kobayashi’s influence and deal directly with county and regional-level elites.

After it became clear that Yoshida villagers no longer accepted Kobayashi Owari no kami as one of their military leaders, Shingen ordered his deputy Oyamada Nobuari to oversee Yoshida directly. Yoshida was thus placed under Nobuari’s direct supervision.

Villagers immediately took advantage of this situation to weaken other Kobayashi prerogatives. The council, called the “more than one hundred persons” (hyaku-yo nin 百

余人) deliberated and sent another appeal to Oyamada headquarters at Yamura later in

1556.84 Once again citing “repeated mistreatment” they leveled charges against the other prominent Kobayashi in Yoshida, Izumi no kami, and demanded an end to their

82 Fujiyoshida-shi shiryō-hen dai sankan, 557. 83 Kobayashi Owari Nyūdō’s 1475 donation to Omuro Sengen Shrine cited above is one such example. Donation registers (kanjinchō 勧進帳) from late medieval Kai indicate that wealthy residents without titles or surnames also gave significant funds, sometimes outstripping individuals of higher status. This reinforces the proposed notion that local landowners like Kobayashi were not autocratic lords of the villages where they lived, but one of a number of wealthy and influential residents. 84 Myōhōji-ki, 45; For analysis of the term hyaku-yo nin, see Nishikawa, Chūsei kōki no kaihatsu, 102.

108 yorioya/yoriko relationship with him as well.85 While this process was underway,

Kobayashi sent a retainer named Danjō 弾正, a warrior, to Yamura as well. Danjō plead

Izumi’s case and offered apologies on his master’s behalf for the rancor between him and Yoshida residents. Discussions must have been intense, for Danjō made the journey to and from Yamura three times in a single day.86 But once again, the Oyamada lord could not render a decision. Perhaps due to his inexperience, or possibly the young

Nobauri was savvy enough to realize he was caught between a rock and hard place. He may have balked at the possibility of alienating either his retainer or the village, and wisely left the decision to his superior.

Kobayashi Izumi then adopted the tactic used previously by Yoshida, and appealed directly to Shingen. He dispatched another retainer, named Kobayashi Bunzō 小林文三

(unknown) to Kōfu. Bunzō remained there from the eighth month until the first month of the following year. During that time, Nobuari issued an apology to Shingen for his failure in resolving the matter. When Bunzō returned from Kōfu, Yoshida villagers were summoned to Yamura once again to hear the Takeda lord’s adjudication. Shingen decided that Kobayashi Izumi could remain at his estate in Matsuyama, however, his role as yorioya to Yoshida was revoked.87 In this way, the Yoshida village council had successfully lobbied the regional elite to remove both Kobayashis from influential local posts.

85 Myōhōji-ki, 44. Although not made explicit in this particular case, Yoshida’s issues with Kobayashi Izumi were almost certainly over competing claims to local resources. 86 Ibid., 45. 87 Ibid.

109

The implications of these decisions for Kobayashi claims were not immediately clear. There was confusion and disagreement over just what their local position now was. Very soon after Bunzō returned with Shingen’s verdict, Kobayashi Izumi sent an envoy to Nobuari in the second month of 1557.88 The intention may have been to appeal the decision, but at this time Nobuari’s officials reiterated that Kobayashi no longer had a recognized role as leader in Yoshida village. His relative Owari no kami was again expressly forbade from building a new estate. In essence, this was an acknowledgement by Oyamada that the territory and resources of Yoshida village belonged to the community collectively – not the wealthy local warriors of the area. That arrangement was tested almost immediately.

In the tenth month of 1557, Yoshida villagers cut timber from a forest called

Miyabayashi 宮林 for construction of riparian works along the Katsura River (see Map

2.5).89 The project was to repair dikes that had been damaged by mudslides.

Miyabayashi had been used by Yoshida village collectively for some time. It turned out to be the very same place that Kobayashi Owari intended to use as a source of building materials for his now disallowed construction project of the previous year. Despite their recent censure, the Kobayashi Izumi lineage attempted to lay claim to this forest in

1557. One of Izumi no kami’s sons, Ichihyōe 一兵衛 (unknown), took action. While the

Yoshida woodcutters labored at Miyabayashi, Ichihyōe and followers fell upon them.

Kobayashi confiscated the villagers’ tools, and had the workers “severely beaten”

88 Ibid. 89 Ibid. Miyabayashi was the contemporaneous name of Mt. Otare 尾垂山, a point discussed further below. The villagers harvested pine, the preferred wood for construction of water works because of its resistance to rot.

110

(nizoku wo sanzan ni tataki 人足ヲ散々二タタキ).90 Yoshida’s response was swift and unequivocal. The village sent the militia to the Kobayashi estate at Funatsu. Armed men surrounded the residence and demanded return of the tools. In other words, Ichihyōe and his cohort were placed under house arrest until they returned Yoshida’s property.

They did not have many options, and so appealed to superiors for help. Ichihyōe managed to dispatch a messenger to his lord Oyamada Nobuari, who was several days away in Shinano on campaign against the Uesugi at the time. Interestingly, Nobuari declined to intervene on the grounds that no wrong had been done (betsu-gi kore naku sōrō 別儀無之候). This was tacit consent for the Yoshida militia’s actions – Ichihyōe should return the tools. Nevertheless, the Kobayashi made a second appeal, this time directly to Shingen in Kōfu. The Takeda lord then dispatched an official to Yoshida and the matter was resolved without further strife. Shingen’s decision “restored the relationship between the Yoshida council and Matsuyama (i.e. the Kobayashi).”91

Contemporary sources do not report how exactly the relationship between Yoshida and Kobayashi was restored at this time. But later documents reveal the terms of the

1557 settlement to have been the following. Miyabayashi forest did indeed belong to the village, but was to be reserved for use in emergencies only.92 Kobayashi was explicitly barred from using Miyabayashi, and instead, the Oyamada were granted permission to harvest a limited number of timbers for construction each year. In other words, the Kobayashi got none of what they requested. The Oyamada gained limited

90 Ibid. 91 Hyakuyonin to Matsuyama to no naka naoshi goza sōrō 百余人ト松山トノ中ナオシ御座候. Ibid., 46. 92 Nishikawa, Chūsei kōki no kaihatsu, 104-05.

111 usage rights to the area, and Yoshida essentially got to keep using Miyabayashi as they had been.

This midcentury dispute shows an in-between point along a spectrum from direct confrontation, which tended towards violence, to a kind of half-procedural, half-direct means of dispute resolution. In this case, both stages of the dispute began with a confrontation. Parties on both sides moved swiftly to back up claims, and both sides had ready military force available. Both sides also appealed to county and regional-level superiors in order to advance their local interests. This represents a very significant development in the evolution of dispute resolutions. Two local claimants leveraged outside elites in order to establish what amounted to ownership of a place. Yoshida had seen only two decades before that county-level lords could act as guarantor of their local claims, even against the lord’s own followers. The 1556-57 dispute demonstrated that that was the case at an even larger, regional level as well. Takeda became a powerful patron for Yoshida. This was one of the first instances of common interest between the very top (Takeda) and very local (Yoshida) in late medieval Kai. That trend grew stronger over the next several decades, resulting in some of the most famous construction projects of the era.93

It was not only a procedural evolution, but a discursive one revealed in this example.

The logic behind dispute resolution became more complex. The general evolution can be characterized as a shift from a dispute as a battle, what Fujiki called “battle disputes”

(kassen sōron 合戦争論) and exemplified by events such as the 1475 attack on

93 See chapter three.

112

Kawaguchi, to a complimentary relationship within the sociopolitical hierarchy.94

Consider the following points from the three disputes so far examined. Case 1 (1475) was unmistakably violent – naked force. Case 2, when Yoshida confronted Watanabe, the village ultimately “won” a decision from Oyamada Nobuari. In case 3 (1556-57) there are several additional concepts at work. Yoshida envoys presented the logic or

“reasonableness” of their claims, and the lord Shingen deemed their position to be

“correct” (tadashii 正シイ).95 They condemned both Kobyashis on the basis of crimes

(hibun 非文) that they had perpetrated against the village. This was not just naked force but a discourse of logic and ownership. And finally, when Shingen rendered the verdict in 1557, it was not understood as a victory for Yoshida, although they did achieve their goals vis-à-vis Kobayashi. This settlement resulted in a “restoration” (naosu ナオス) of the relationship between the Kobayashi Izumi lineage and Yoshida villagers.96 In other words, there was now a clear idea of how the relationship between these two entities should function. Furthermore, it established a mutually beneficial relationship between small-time Yoshida and the regional magnate. The village got help resolving an internal problem, and the Takeda quashed a potentially disruptive conflict. Still it took some time for these new arrangements to be firmly ensconced, as indicated in the next example.

94 Fujiki, Sengoku no sahō, 14. 95 Myōhōji-ki, 47. 96 Ibid.

113

E. Case 4: Local “Ownership” Confirmed (Yoshida vs. Kobayashi, 1559)

One of the major implications of the 1556-57 dispute was the de facto investiture of

Yoshida as owners of Miyabayashi forest. Villagers would refer back to the settlements of the mid sixteenth century generations later as precedent establishing their ownership.97 A second dispute over Miyabayashi in 1559 solidified the forest’s status effectively as Yoshida communal property. The dispute began when the village was ordered by Oyamada Nobuari (Yasaburō) to contribute to a construction project.

Yoshida was requisitioned to supply timbers for “river maintenance” (kawayoke 川除), and were expected to harvest these materials from Miyabayashi.98 This order from

Oyamada was issued to the Yoshida “administrative group” (hōri-shū 祝師衆・法理衆), who had exclusive rights to the forest.99 The Yoshida administrative group was none other than the village council, also called the “group of one hundred” (hyakuyo nin) or simply the “Yoshida group” (Yoshida-shū 吉田衆).100 The basic structure revealed by

Oyamada’s 1559 order are therefore the following: 1) Yoshida villagers were expected to participate in river infrastructure projects by contributing materials (and labor, as specified in other documents); 2) responsibility for organizing that participation was delegated to the authority of the village council; 3) contributions of material resources were expected to come from the village’s own stores, in this case the Yoshida-held

Miyabayashi; and 4) the village council was in charge of administering communal

97 Fujiyoshida-shi dai sankan, 557. 98 Myōhōji-ki, 46. Kawayoke was a general term for riparian construction which include such projects as: river dredging, construction of dams, sluices, dikes, water breaks, or reroutes, and seasonal clearing of debris. 99 For analysis of the hōri-shū, see Nishikawa, Chūsei kōki no kaihatsu, 102–4. 100 Ibid., 102; Osano, “Takeda ryōgoku no dogōsō,” 338–39; Shibatsuji Shunroku, “Sengoku-ki no suiri kangai to kaihatsu,” Minshū-shi kenkyū 11 (1973): 85.

114 property. In other words, this body, which was comprised of adult residents of Yoshida, effectively governed Yoshida.

Thus far, the situation seems fairly straightforward, but it resulted in a dispute over forest rights because the logic of ownership was still in flux. We know from Shingen’s settlement of the 1556-57 dispute that Miyabayashi was supposed to be reserved for use in emergencies by Yoshida. That appears to have been the case in 1559. Heavy snow the previous winter contributed to a very wet spring, with above-normal levels of snow melt, and consequently damaging mudslides.101 Oyamada’s kawayoke order was for construction of a levee to protect the tutelary shrine, Omuro Sengen (see Map 2.3). Once this project was underway, he appointed Kobayashi Owari no kami as overseer (bugyō

奉行). Although Kobayashi had just been ousted as a local military leader, he was now in a position to serve as his lord’s (Nobuari Yasaburō) local deputy in Yoshida.102 This role was in his capacity as an official (bugyō) of Oyamada not as a local leader. In other words, the relationship between Kobayashi and the rest of the village was beginning to change. The rank and file in Yoshida were using external allies (Oyamanda and Takeda), to curtail the power of a wealthy resident who had run afoul of the larger community.

The events of the 1559 dispute reveal an increasing distance between the Kobayashi and the rest of Yoshida, ultimately leading to their exclusion. 103

101 Myōhōji-ki, 47. 102 Yoshida had been placed under the supervision of Nobuari by his lord, Shingen, after the 1556-57 disputes. 103 This is the “pushing” of small-samurai out of local areas that Birt has described. His focus is administrative and political, but it is important to note here that the impetus for that push was conflicts over local resources. Birt, “Samurai in Passage, 398–99.

115

Although it was Oyamada Nobuari who “ordered” new flood works in 1559, this project was certainly at the behest of local Yoshida leaders. New construction was designed to protect Omuro Sengen Shrine. The Shrine was important communal property in Yoshida. It mattered to them, not to Nobuari. Nor would the Oyamada lord have any real knowledge about how or why such a project should be undertaken, thus the appointment of local resident Kobayashi Owari as overseer. Nobuari’s construction order was therefore a tool local leaders used to organize participation. It was not in itself the driving force behind the project.104 And, in contrast to the 1556-57 case, it was for collective property, not a new Kobayashi estate.

Kobayashi Owari seems to have accepted his new role and the restrictions placed upon him. But Kobayashi Izumi was still not ready to give up claim to various local resources in Yoshida. As they mobilized to carry out Oyamada Nobuari’s construction order, Izumi expressed to Yoshida villagers that they should not enter Miyabayashi to cut timbers because rights to the forest belonged to him.105 But the villagers ignored this and continued their work. Izumi then pressed his case no less than three times with

Nobuari himself.106 Not only was this claim denied, but Izumi was ordered to return property he had taken from Yoshida some ten years prior. This order was issued to settle an old grudge over a water dispute in 1550. That year, with Kobayashi Owari no kami acting as principle sponsor, the Yoshida village council had organized construction

104 Sasamoto Shōji has argued for the reconceptualization of many medieval historical documents along these lines. Surviving historical material from this era is almost exclusively products of elites. Edicts, laws, and writs of investiture discursively appear very one-sided, reflective as they are of the sociopolitical hierarchy. However, Sasamoto repeatedly points out that the vast majority of these documents were responses to subordinates’ requests (which either have tended not to survive or were never themselves documented). 105 Myōhōji-ki, 47. 106 Ibid.

116 for a new levee that would redirect the waters around Owari’s estate.107 The Izumi

Kobayashi were opposed to this project, and the aforementioned Ichihyōe physically stopped work on the dike by stealing shovels and hoes from the Yoshida village communal store (hyakuyonin-shū no tokoro 百余人衆所) and holding them as collateral

(shichimotsu 質物).108 Confiscation of the community’s tools proved to be an effective way to stop work. The fact that even after ten years, Yoshida villagers sought the return of their tools indicates how valuable and difficult to replace these implements were. But in the end, Izumi’s last-ditch effort to hold on to some claim to Miyabayashi failed. It ultimately reinforced Yoshida’s exclusive rights to the forest. Crucially, this settlement was laid in place via non-violent procedure.

These four cases from late medieval Yoshida highlight the fact that resource disputes were the primary forms of conflict in civil war era Japan for most people. The emerging regularization of dispute resolution originated from, at least in the local context, resource disputes. These are the key conflicts for which local sources record with greatest frequency and detail. That is not to say there were not other kinds of disputes. But resource disputes have left the largest footprint in historical materials from late medieval Yoshida. Resource disputes continued into the early modern period and beyond. But after widespread pacification in at the turn of the seventeenth century,

107 This was prior to Kobayashi Owari’s rejection as yorioya of Yoshida. Unfortunately, we do not have many physical details of this project, making it difficult to various residents’ reasons for support or opposition. It appears to have been backed by most of the community. 108 Myōhōji-ki, 47; Fujiyoshida-shi shiryō-hen dai nikan, 386. Note that taking tools was Ichihyōe’s tactic again in 1556, to stop Yoshida villagers from cutting wood at Miyabayashi.

117 resource disputes proceeded very differently than they had during the latter decades of the civil war era.

F. Case 5: Judicial Resolution (Yoshida vs. Kurechi and Onuma, 1635)

In Kan’ei 寛永 2 (1635), Yoshida disputed competing claims to Mt. Otare from the nearby villages of Kurechi 暮地 and Onuma 小沼 (see Map 2.6). This dispute was fully procedural, consisting of a lawsuit and judicial examination. Village militias were not involved and no violence took place. The dispute began when Kurechi and Onuma endeavored to use Mt. Otare for slash and burn argriculture (sasubatake さす畑).109 Mt.

Otare was in fact Miyabayashi, still an important resource cache for Yoshida after many generations. And it continued to be so, referenced in documents from 1635, 1759, and

1806 as Yoshida’s source of timber for riparian construction projects.110 As such,

Yoshida took steps to vigorously oppose this encroachment on their territory. But instead of sending out the militia, they filed suit. The village council, called sō-hyakusho

総百姓 at this time, submitted a petition to the local magistrate (bugyō) under Akimoto

Yasutomo 秋元康朝 (1580-1642), a Tokugawa retainer entrusted with Yamura 谷村 domain at that time.111 Yamura was essentially the old Oyamada territory, the Gunnai

109 Fujiyoshida-shi shiryō-hen dai sankan, 556. 110 Precisely as Miyabayashi had been defined in the previous era, suggesting with near certainty that there were the same place. See Nishikawa, Chūsei kōki no kaihatsu, 103-04, 107. 111 Akimoto’s official in Yoshida is unnamed.

118

Kurechi Onuma Mt. Otare

Shimo no mizu

Shinkura

Yoshida

Map 2.4: Key Sites of the 1635 Dispute

section of Kai Province, and Akimoto’s magistrate was based at the Funatsu checkpoint, which the Kobayashi had once controlled.112 In their petition, Yoshida argued in no uncertain terms that the forest belonged to them. The other villages’ claims to this area were completely unjustified. Yoshida leaders based these claims on three primary factors. Foremost was that the forest had been designated as Yoshida’s source for miscellaneous raw materials (karishiki 刈敷) and as such they had paid a portion of

112 Nishikawa, Chūsei kōki no kaihatsu, 106.

119 these materials as tax (karishiki nengu 刈敷年貢) since the early 1600s.113 Furthermore, other villages had been denied usage and development rights to the forest on Mt. Otare in the past. And finally, the area had been designated as under Yoshida’s responsibility for previous road and river construction projects.

Yoshida officials began their appeal by citing tax obligations as a kind of proxy for ownership. According to the Yoshida suit, the village had paid a materials tax on Mt.

Otare for some thirty-two years.114 This entailed collection of undergrowth and other plant materials to be used as fodder, fertilizer, and raw materials for goods such as thatch and rope. Yoshida paid this tax on Mt. Otare because it was where the village gathered these materials for their own use. The fact that Yoshida residents were the ones using this place and paying taxes from it meant it belonged to them. Yoshida explicitly denied that Kurechi or Onuma have any justifiable claim to Mt. Otare. The logic at work here is straightforward but significant: Yoshida paid taxes on Mt. Otare because they are the ones that use it, and therefore, it belongs to them.

The next points in the suit proceeded roughly chronologically, itemizing the village’s claim to Mt. Otare. During the time of master Saemon no suke 左衛門佐様 (Asano

Ujishige 浅野氏重, d. 1619), Shinkura village 新倉村 was denied rights to develop a parcel of land between Mt. Otare and an area called Shimo no mizu, along the banks of the Katsura (see Map 2.4).115 Yoshida had prevented this by appealing to Ujishige’s

113 Fujiyoshida-shi shiryō-hen dai sankan, 556. 114 Ibid. 115 Ibid. The events described must have taken place in the 1590s. Ujishige was enfeoffed in Tsuru county, former Oyamada territory, by his lord Asano Nagamasa 浅野長政 (1547-1611). Nagamasa was entrusted with the former Takeda domain by Hideyoshi in 1593. After the Takeda fell in 1582,

120 magistrate at Funatsu, one Ishizumi Zen’emon 石墨善右衛門 (unknown). Ishizumi endorsed Yoshida’s claim to this spot of land, asserting that it was attached to Mt. Otare.

As evidence of that decision, the 1635 appeal points to the fact that Yoshida paid the materials tax on this plot of land to Ishizumi’s officials Kawaguchi Kakuzaemon no jō 川

口覚左衛門尉 (unknown) and Itō Matabei 伊藤又兵衛 (unknown).116 Yoshida’s appeal then describes the assignment of construction duty, which further supports their assertion that this contested area belonged to them. During the era of master Tosa no kami 土佐守様 (Tori’i Naritsugi 鳥居成次, 1570-1631), there was major river dredging of the Katsura in Gen’na 6 元和 (1620).117 Part of this project involved building a road that was to run parallel to the river. As per their corvee duties (fushin-yaku 普請役),

Yoshida was responsible for the section of the road “above” (upstream – in this case southwest) Shimo no mizu.118 In other words, Naritsugi’s order acknowledged that this territory belonged to Yoshida.

The most significant parts of Yoshida’s 1635 appeal are the fourth through sixth items in the second half of the document, in which the villagers cite decisions from some of the sixteenth century disputes examined above. First, Yoshida asserts that during the time of Owari Nyūdō Lower Yoshida, adjacent to Shimo no mizu, had

Tokugawa Ieyasu took over the domain, but was transferred to the former Go-Hōjō territory following their demise in 1590. The domain changed hands a number of times in Hideyoshi’s era, including to the Asano. After Sekigahara, the Tokugawa and their close relatives reclaimed direct control of Kai Province. 116 Ibid. This Kawaguchi was not related to the lineage eliminated in 1475. 117 Naritsugi was given Yamura domain in 1600 and held it until his death in 1631. His son Tadafusa 忠房 (1606-1637) succeeded him but was attaindered by the bakufu in 1633 and Yamura passed to the Akimoto. The old Takeda seat at Kōfu was the castle town of a larger domain (Kōfu-han 甲府藩) belonging to Naritsugi’s lord Tokugawa Tadanaga 徳川忠長 (1606-1634) at the time. 118 Fujiyoshida-shi shiryō-hen dai sankan, 556.

121 developed new dry fields (tahata たはた) and used materials from Mt. Otare.119 This

Owari Nyūdō almost certainly refers to Kobayashi Dōkō, who sponsored the 1504 infrastructure projects discussed above.120 The appeal continues that “for four generations, from the time of master Keisan to Nobushige” (keisan-sama yori nobushige made yondai けいさんさまゟのふしけまて四代), and afterwards, Yoshida had cut lumber and undergrowth at Mt. Otare.121 The assertion here is quite clear: Mt. Otare belongs to Yoshida and it has for the past century. The four generations of Oyamada described here span nearly the entire 1500s. Keisan refers to the elder Oyamada

Nobuari, Etchū no kami, and Nobushige succeeded his brother, the third Nobuari, in

1565. In other words, the mid sixteenth century settlements to the Yoshida v. Kobayashi disputes established Yoshida’s claim and continued to be the basis for management of

Mt. Otare resources.

Responses to Yoshida’s 1635 appeal have not survived in the historical record.

However, we know from subsequent documents that Mt. Otare continued to be attached to Yoshida village as late as the early nineteenth century.122 That relationship was maintained by legal proceedings such as the one examined above, which were justified by arrangements that had been worked out in the sixteenth century. Decisions rendered even as early as the first decade of the sixteenth century served as the basis of property claims into the nineteenth century. This fact highlights a major assertion of the

119 Ibid. In other words, Mt. Otare was the source of fertilizer for these fields. 120 It could be referring to Dōkō’s father Shōki, who also had contributions to local infrastructure such as his 1475 donation to Omuro Sengen Shrine. See note 47. 121 Fujiyoshida-shi shiryō-hen dai sankan, 556–57. 122 These documents are published in Fujiyoshida-shi kyōiku iinkai, ed., Fujiyoshida-shi shi shiryō sōsho 4, mura meisaichō (Fujiyoshida-shi: Fujiyoshida-shi kyōiku iinkai, 1988); See Nishikawa, Chūsei kōki no kaihatsu, 102-04 for a summary.

122 current project. Local settlements regarding resource endured as the basis of local administration from the era of Sengoku daimyo well into the early modern era. These practices were solidified in Kai around midcentury, and cases 2, 3 and 4 in particular indicate a transitional moment when the new terms of use, rights, and ownership were contested. As the 1635 example shows, subsequent lords endorsed these decisions.

They became standard practice and a kind of common law. Although there were changes of many kinds to local administration in the early modern era, the basic structure had been put in place by the mid sixteenth century.

At issue in the 1635 case were usage rights to a forest, and Yoshida’s response indicated that they had enjoyed effective ownership of this place for generations. There was no equivocation, Kurechi and Onuma’s claims were expressly denied, and Yoshida’s appeal laid out how and why no one else had used Mt. Otare for over a century. Yoshida did not own this land in the modern sense, but determined who entered the forest and who was allowed to take what from there. This amounted to ownership for all intents and purposes in premodern Japan.

The lack of violence in the 1635 dispute is a striking difference from examples during the civil war era. Local militias do not appear to have been involved in any way.

This dispute began and ended procedurally. At the start, Kurechi and Onuma requested permission to use the disputed land. Yoshida’s response was to challenge that request via legal suit. This represents quite a departure from the sixteenth century examples we have examined. During the civil war, disputing parties’ initial response was to send out the village militia or otherwise directly confront their opposition. We saw this

123 numerous times with Yoshida, Watanabe, and Kobayashi. But here, conflict resolution reflects a dominant rule of Tokugawa era lordship, which was to respect all preexisting arrangements that had been in place for twenty years or more.123

G. Conclusions: Civil War-era Disputes in Context

This chapter has attempted to describe an evolution in the how local resource disputes proceeded and how they were resolved in late medieval Kai. The focus on

Yoshida has demonstrated a trajectory from disputes as direct and violent, to procedural and institutionally mediated. Several factors contributed to this fundamental change. From the perspective of the local, mediation by elites was not an obvious solution in the midst of the civil war. “Basic government services,” to borrow

Adolphson’s phrase, had only marginally functioned throughout most of the medieval era. By the time of the case studies presented here, effective institutions had not existed in Kai Province for almost a century. Yoshida itself was at the intersection of several competing warlord territories. How could they know who would have the capacity to assist them in asserting their local claims, or what the terms of that assistance would be? There was in fact no way for them to know aside from experience. It was for that reason crucial that county and regional-level elites demonstrated the wherewithal to act as mediators, but also willingness to endorse local claims. Their role as guarantor was the reason locals appealed to them in the first place. Most scholars agree that that

123 Luke Shepherd Roberts, Performing the Great Peace: Political Space and Open Secrets in Tokugawa Japan (Honolulu: University of Hawai'i Press, 2012), 118, 120.

124 was the basic process by which elite warlords gained the allegiance of lower ranking warriors as well. Daimyo gained followers by endorsing their property claims. Takeda confirmation documents issued to small-time retainers abound, and these arrangements were carried over almost unchanged under Tokugawa administration and beyond. Within the warrior class, key stakes included the terms of military service, rank, economic incentives, and geopolitical strategy. For local “ordinary” people, claims were centered upon resources. At stake were the variegated livelihoods dependent upon those resources, and the autonomy to manage the local environment as residents saw fit. Late medieval and early modern rulers primarily left those issues to locals, so long as basic obligations were met. For both the very elite and very local people, this arrangement offered advantages of stability over what had prevailed during the civil war. It was, in the words of historian Wayne Farris, “a more regular, palatable, and bearable form of macroparasitism.”124

The above examples show why local residents in a place such as Yoshida would ultimately have an incentive to accept daimyo authority as legitimate. It served their interests in what were to them the most pressing issues of the day: local conflicts. These disputes also show that the most contentious processes in late medieval Yoshida were not submission to a territorial lord, but fights within the community itself. The Takeda helped resolve some problems, ultimately by accommodating local preferences. From the perspective of the Takeda, this was the most expedient, effective means of incorporating territory. It also helped limit the independence of small-time retainers

124 Farris, Japan’s Medieval Population, 266.

125 such as Kobayashi, who could be sources of unrest. All of these things contributed to the warlord’s primary goal: ensuring local stability. The Takeda therefore based their judgments of local disputes on what would be the most effective way to keep the peace.

Delegation of much of the authority to manage local space to local themselves turned out to be the most effective option. The needs of regionally-powerful Takeda and those of small-time Yoshida actually dovetailed in a way that reestablished order.

The disputes in Yoshida show how premodern disputes and litigation in general evolved from the late medieval to early modern eras. In the late civil war, the ability to resolve disputes was a crucial, valued kind of sociopolitical capital. The risky, unpredictable system of “self-redress” gradually morphed into a practice of delegating mediation into the hands of territorial lords. As many have pointed out, this was the primary means by which several Sengoku daimyo justified their rule. The Mōri 毛利, for instance, were quite explicit in stating that their authority was legitimate based upon their ability to mediate disputes and maintain order in the domain.125 Mōri were especially keen to make this case because they had a legitimacy problem, supplanting as they did the “rightful” lords of Aki Province 安芸国 (modern Hiroshima Prefecture), the

Ōuchi 大内. But even warlord lineages with legitimate pedigrees, such as the Takeda, created new ways of justifying their rule. Takeda Shingen issued his famous law code

Kōshū hattō no shidai 甲州法度次代 as a way to institutionalize the power relationships he had worked to establish in Kai. These collectively became the basis of the discourse of “public good” (kōgi 公儀), the idea that subjects of the domain shared basic common

125 Farris, Japan’s Medieval Population, 224; Eason, “The Culture of Disputes,” 76.

126 interests (such as the desire for stability). The examples from Yoshida show how this discourse was formed, and how it actually operated. For the vast majority, any notion of common interest was closely tied to local resources and communal management of the environment.

The issue of local autonomy and its role in the transition from medieval to early modern has generally been portrayed as a kind of perk afforded to communities by the new ruling class. Tight supervision by on-site estate officials and subsequently independent landlords was replaced by a loose oversight from a centrally-administered domainal government. That does not appear to have been the case in Yoshida. If anything, the village surrendered some of their autonomy, like the ability to deal with disputes directly using the militia. But they gained tools for more effective, smoother local administration. Demonstrated over the course of the sixteenth century, cooperation with the Takeda actually helped the community achieve their local goals.

But this was a new kind of subjugation. It was one that favored stability above all else.

The archipelago had experienced a century and a half of civil war. While it did afford a high degree of local autonomy, that autonomy was restricted in a specific framework that had not existed during the civil war. Takagi Shōsaku describes the process as the confining of autonomy within “general peace” (sōbuji 総無事).126 Anything that could upset the general peace was firmly disallowed, on the basis that it would upset the peace. Hideyoshi most famously expressed this philosophy in his peace edicts of the

126 Takagi goes on to say that within that framework, however, “…it was not only desirable but even essential for individuals and communities to exercise autonomy within the limits allowed by the [early modern] state, while paying due heed to the will of the state. In other words, the early modern state relied upon local society to more or less function on its own. Takagi, “Hideyoshi’s Peace and the Transformation of the Bushi Class,” 66.

127

1590s. In their assessments of these edicts, Takagi, Fujiki, and others have argued that this was what established a new sociopolitical framework and opened the door for the developments of the early modern era.127 This is a classic “top-down” assessment of the

Sengoku and early Edo periods. Takagi even goes as far as to say that Hideyoshi’s military might was both the “ends and means” which established order and enforced peace.128

Takagi and Fujiki’s conclusions regarding autonomy within a specified system of constraints fails to elucidate how that framework was formed in the first place. It is unrealistic to think that military force was both the ends and means of societal transition in a society that had been at war for over a century. Everyone had force at their disposal, even on a very small scale like Yoshida. The discourse of kōgi and sōbuji were not imposed upon a blank slate. Hideyoshi himself adapted policies of various regional lords to a larger scale, not the other way around.129 And underpinning those domains’ effective control of territory were the local arrangements such as those examined here.

The top-down analysis of the medieval to early modern transition has been modified in the years since Takagi and Fujiki were writing, yet very few Western scholars have offered serious critiques.130 Those that have tend to frame their analysis in terms of elite and local interactions and conflicts. Wayne Farris and Philip Brown, for

127 Takagi, “Hideyoshi’s Peace”; Fujiki, Toyotomi heiwarei. 128 Takagi, “Hideyoshi’s Peace,” 60. 129 Eason, “The Culture of Disputes in Early Modern Japan, 1550-1700,” 122. 130 This is partially a function of the low number of English-language studies of the late medieval period.

128 instance, both emphasize that new political arrangements were a give and take between rulers and ruled.131 The majority of contemporary Japanese scholars agree with that conclusion, as do the findings here.132 But hopefully this study will help us move beyond a dichotomy between rulers and ruled. By analyzing local environmental management, we can see how much civil war era conflict was by necessity between locals, and rulers had to accommodate to their local solutions. Regional lords such as the Takeda were remote, distant, though not unimportant, parties to such conflicts. Examining the local environment can reveal the complexity of sociopolitical relationships within the local community, at the county level, regionally, and on larger scales.

My hope is that this chapter has demonstrated how local resource disputes played a formative role in ending Japan’s long age of civil war and the reestablishment of stability. The early modern state, for all its bluster, was in fact extremely flexible.

That flexibility was routinely put into practice in the context of local administration.

When domainal governments did intervene in early modern resource disputes, though their decisions could be punitive, they gave a high degree of deference to local practice.

It was precisely this flexibility which explains the remarkable durability of the

Tokugawa order.133 Ecologically, localized environmental management preserved favorable environmental conditions that kept local areas viable for residents for many generations.134 When these practices were subjected to centralized, national control,

131 Farris, Japan’s Medieval Population; Brown, Central Authority and Local Autonomy. 132 That trend could generally be said to have begun in the 1980s with the work of Amino Yoshihiko. 133 Luke Shepherd Roberts, Performing the Great Peace: Political Space and Open Secrets in Tokugawa Japan (Honolulu: University of Hawai'i Press, 2012). 134 Favorable from the perspective of the local human community. For examples of how this kind of local management maintained key ecological relationships locales around Lake Biwa, see Sano, Chū- kinsei no sonraku to suihen no kankyōshi, 213–15; Sano Shizuyo, “Traditional Use of Resources and

129 these relationships dissolved and the impact on the local environment was less than favorable. Late medieval daimyo did not expand their influence by taking direct control of vast resources.135 It was rather their ability to effectively manage disputes over resources which solidified their position. And that continued to be the foundation for the stability of the early modern state.

Management of Littoral Environments at Lake Biwa,” in Batten and Brown, ed., Environment and Society, 90. 135 This point is discussed further in chapter 4.

130

III. Subjugation to Stability: Building Complimentary Interests across Socioeconomic Levels

The previous chapter demonstrated how locals used elite influence to advance their own parochial interests, often vis-à-vis neighbors competing for the same limited resources. A central contention is that these contests represented an extremely consequential type of conflict during the civil war era for the majority of residents. Reliable resolution of endemic rivalries over local resources had a significant impact upon overall stability. These solutions became possible because the interests of the very elite and very local gradually became complimentary. This chapter explores how mutually beneficial relationships were forged across the socioeconomic spectrum in late medieval Kai.

After decades of upheaval, local residents and a new class of regional, as opposed to central, elites came to rely upon one another. Local resource disputes played an important role in establishing that relationship. This reconfigured hierarchy gradually solidified into a well-defined system of practices. Regularization in turn led to a greater vested interest in the maintenance of the emerging sociopolitical order. And, as that vested interest broadened socioeconomically, the new order became ever more entrenched. The previous chapter showed how that process gained momentum in sixteenth century Yoshida. It has generally been analyzed from the perspective of daimyo state-building. Certainly it is true that local people in Kai ultimately submitted to Takeda authority and were integrated into Takeda administrative structures.

131

However, these structures were formed collectively. They placed responsibilities, privileges, and limitations on both inferiors and superiors in the sociopolitical hierarchy. Everyone became subjected to structures that functioned above all else to maintain stability. That basic arrangement was finally achieved in Kai circa 1560. It endured throughout Tokugawa rule of the province and into the early modern era.

Complimentary interests of socioeconomic groups were limited in scope but had a broad range of consequences. At the local level, critical issues involved comparatively basic, although difficult to implement, political functions. As we have seen, local residents wished to have their claims endorsed by higher authorities, and they wanted those authorities to be willing and able to protect those claims when challenged. To achieve those goals, residents needed a mechanism for resolving disputes, in particular those against fellow local rivals. All kinds of people, regardless of occupation, stood to benefit from greater certainty that their claims could be enforced. Once that certainty could be demonstrated, integration into a hierarchical relationship was generally preferable to the risk of going it alone. At the very elite level, vested interest in maintaining such hierarchical relationships was predicated on their ability to remain at the top of said hierarchy. Doing so required integration of geopolitical and economic strategy in such a way that their claims to authority were more effective, thus more “legitimate,” than that of competitors. Daimyo interest can therefore be simplified into three categories: maintenance of internal stability, command of (or rights to) a portion of local productivity, and effective neutralization of external threats. The most important needs of both local people and regional elites do not necessarily conflict, and in fact dovetailed in important ways. Because Takeda

132 strategy operated on a provincial scale, details of “on the ground” administration of local areas was much less consequential overall. In other words, usually the best way to administer territory was to confirm claims that were already in place. These had developed organically over decades of civil war, and institutionalization of local practice was what residents desired. The only major sticking points were twofold: 1) how to minimize disruption when resolving competing local claims; and 2) how to adjust service obligations in a way that benefitted the Takeda while at the same time kept most local practices intact. We have examined how the former challenge was navigated in a number of cases from late medieval Yoshida. Examination of Takeda economic policies throughout the rest of the domain will show how the latter was achieved. These policies are examined from three perspectives: incorporation – essentially the process of confirmation itself, service – the establishment of obligations and their terms, and infrastructure – projects that brought local interest, elite strategy, obligations, and property claims together for a specific goal.

A. From Province to Country: Incorporation

Civil war era regional polities differed significantly from previous and subsequent entities. Estate and provincial land (shōen-kokugaryō 荘園国衙領) connected central and provincial elites in a web of overlapping rights, while early modern domains (kuni

国) were subject to significant oversight, especially of the daimyo personally.1

1 Early modern domains are usually referred to as han 藩 in scholarly literature. Luke Roberts has argued that use of this term obscures the ways contemporaries understood the Tokugawa-era political hierarchy. Domains were largely autonomous within a very specific framework, which

133

Warring States domains, on the other hand, were independent. For that reason, scholars have employed the term ryōgoku 領国 (domain) for the territorial states of the late medieval era in recognition of their qualitative difference. But here too, the use of the word “domain” implies that these territories were parts of a larger whole.2

While contemporaries certainly acknowledged interconnections with other regions, especially adjacent ones, there is little indication of that residents viewed Kai

Province as part of a larger, coherent whole in the late medieval period. The place itself was one of many provinces (kuni or shū 州), as defined by ancient borders established by the court. From the local perspective, the big change in late sixteenth century Kai was not that the province became integrated into a newly “unified” realm

(tenka 天下), but that the Takeda domain (bunkoku 分国) became the functional regional political entity.3 Kai Province had long been a geographic denomination, absent any real political significance. But by the mid 1520s the Takeda domain, which ultimately encompassed Kai, most of Shinano, and parts of several other provinces, was a functional political unit. This was more than a shift of definition, it affected people’s livelihoods.

While we usually think of daimyo rule in terms of territorial control, direct oversight or administration of local areas by Warring States daimyo was extremely limited. This was definitely true in the Takeda case. They exerted authority indirectly

helps explain some of the seeming inconsistencies or weaknesses of the Tokugawa government. See Roberts, Performing the Great Peace. 2 Which also justifies the narrative of national “unification” at the end of the sixteenth century. 3 Tenka was the term Oda Nobunaga used to invoke all of Japan, a diverse geopolitical space he endeavored to bring under his influence in the so-called “unification period” (Azuchi-Momoyama 安土桃山, ca. 1560-1590) of the civil war era.

134 by controlling the local power holders and levying them with various service obligations. As in other daimyo domains, the real mechanism of territorial rule was the terms of those obligations.4 For the Takeda, the most important was between those who owed military service (gun’yaku 軍役), and those who did not.5 The military service group (gun’yakushū 軍役衆) formed the core of Takeda’s hereditary servants and official agents. Non-military livelihoods for the most part were not classified into specific “groups,” (shū 衆) although some occupations were.6 But because these individuals did not owe the Takeda hereditary military service, their relationship with the lord was fundamentally different. Military retainers were nominally awarded property and tax exempt status in exchange for their service. All others nominally owed taxes, in the form of cash, goods, or specialized services. The

Takeda based political and economic policies upon this basic distinction, but its actual implementation was far more complex.

Before examining that complexity, consider why any occupational group would enter into such a relationship with the Takeda in the first place. The Takeda house was shugo daimyo under the Muromachi regime, and although that pedigree did not immunize them from the upheavals of the fifteenth century, it did afford the lineage a certain amount of wealth and prestige. However, other warrior houses with varying

4 What was new about this sociopolitical arrangement was that these obligations could only be established within a confined, contiguous geographic area in which the daimyo or his allies could muster effective (i.e. military) influence. That differed in important ways from the traditional order described in chapter 1. 5 This distinction can be considered a precursor to the separation of professional warriors from other social classes with the official status of samurai. 6 Including gold miners, lumberjacks, carpenters, and some merchants. In cases where non-military personnel are classified as a “group,” it is geographically based. For example, the “western sea group” (nishi no kai-shū 西之海衆) of shippers in Kai’s Koma County.

135 degrees of the same competed for supremacy in Kai for most of the fifteenth century.

The Takeda were but one of many, if the most recognizable of the bunch. When

Takeda Nobutora came to power as a teenager in 1508, the Takeda house began to separate themselves from competitors. Nobutora turned out to be a very successful

(and lucky) military leader. By the early 1520s he had militarily bested his major rivals in Kai. That was no small feat, but did not in itself guarantee Takeda political supremacy. His mechanism for cementing Takeda rule followed a formula that would be expanded upon, but largely unchanged, by his successors for the remainder of the sixteenth century: confirmation in exchange for service.7

Confirmation of rights or privileges in exchange for service is most clearly evident in the lord-retainer relationship of professional warriors. Some of the earliest and most explicit manifestations of this arrangement in Kai come from Takeda-issued documents during Nobutora’s reign. Kōyō nikki 甲陽日記, the diary of Takeda retainer

Komai Kōhakusai 駒井高白斎 (d. 1563?), provides some insight into the substantive shift that took place in 1520s Kai. After decades of internecine struggle, Takeda

Nobutora achieved a series of victories that culminated in Tai’ei 大永 1 (1521). The following year, 1522, according to Kōhakusai, all the warrior lineages of Kai province came to Shōhōji 正法寺 at Nobutora’s behest to receive confirmation letters

7 I should note here that Takeda Nobutora was by no means the innovator of this arrangement. Although many of the details differed, it was not a qualitatively different than what other warlords worked out within their territories. In terms of its genesis, it is beyond the scope of this project. I suspect that the basic “confirmation for service” relationship between superiors and inferiors did not have a single point of origin and was drawn from a variety of precedents and longstanding cultural conventions.

136

(gohangata 御判形).8 This was an extremely significant event. These documents acknowledged local warrior claims to property and specified their obligation of military service to the Takeda in perpetuity. On the surface, this might appear to be terms imposed by a conqueror upon the vanquished. More important, though, is the implicit acknowledgement of Nobutora’s role as guarantor of retainer lineages’ property.

There were practical reasons why other warrior houses would have benefitted from a service relationship with the Takeda. Several overt manifestations of Takeda power took place in the few years before Nobutora handed out confirmation letters at

Shōhōji in 1522. These began in 1519, when Nobutora moved his headquarters from

Isawa 石和 to the centrally located Kōfu.9 Located in the middle of the largest plain in the province, Kōfu was a much more favorable location. It boasted the most productive agricultural land, and had ample water from the two large rivers (the

Kamanashi 釜無 and Arakawa 荒川) which flow on either side of the Kōfu basin.

Between these waterways, Nobutora built an ample villa at Tsutsujigasaki. In 1521 he took a kind of victory tour around Kai, visiting the most important regional religious sites with a select group of followers. They first went to Kuonji 久遠寺, the head temple of the Nichiren sect, located at Mt. Minobu 身延山 in south central Kai.10

8 Kōyō nikki in Yamanashi-ken shihensan iinkai, ed. Yamanashi kenshi shiryō-hen 6: chūsei 3 jō, kennai kiroku. Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 1999, 84. Shōhōji was a temple that once stood along the ritsuryō-era “Kai road” (kai-ji 甲斐路) near the Kurokoma checkpoint (黒駒関), which was the entrance to the province from the tōkaidō. The area is now part of Fuefuki City 笛吹市. 9 Isawa is now Isawa-chō 石和町 of Fuefuki City. To the southeast of Kōfu, it is in the same area where Shōhōji existed. 10 Kōyō nikki, 84; Myōhōji-ki, 19.

137

Nobutora’s entourage then traveled to Mt. Fuji, where the Takeda lord himself ascended the summit.11 After visiting the two most prestigious mountains in southern

Kai, the pilgrimage continued at Zenkōji 善光寺 in Shinano.12

The message behind the construction of the new villa in Kōfu and Nobutora’s extended pilgrimage was clear: Takeda influence extended across the entire province.

This does not mean that Nobutora was the universally recognized ruler of a unified territorial state. On the contrary, his position was challenged repeatedly from within

Kai throughout the 1520s and 1530s. But it did indicate to potential rivals (and allies) the capabilities that Nobutora had at his disposal. In other words, many warrior lineages realized that it was in their best interest to secure Takeda backing, and that of course meant agreeing to the terms of a lord-retainer hierarchy. The new capital and Nobutora’s pilgrimage were not so much a victory tour as a recruitment tool.

They demonstrated what the Takeda could (potentially) offer lesser lineages. And indeed, Takeda vassals gained a measure of security over their most pressing concern, maintenance of claims, by their submission.

The terms of confirmation letters show how both lord and retainer benefitted from the hierarchical relationship established. These documents, known generally as andōjō 安堵状 (lit. “writ of recognition”), always acknowledge a retainer’s claim to a

11 Myōhōji-ki, 19. 12 Kōyō nikki, 84. Zenkōji was and remains the most prominent temple in the Kai-Shinano-Echigo region (kō-shin-etsu 甲信越) of Japan. It is one of the oldest temples in the country, so old in fact that it predates the arrival of sectarian differences and therefore belongs to none. Zenkōji’s principal image is the Ikkō Sanzon Amida Nyorai 一光三尊阿弥陀如来, supposedly the oldest Buddha statue in Japan. It is a secret Buddha (hibutsu 秘仏), a replica of which is displayed to the public only once every seven years, to much fanfare. See Sasamoto Shōji, Zenkōji no fushigi to densetsu: shinkō no rekishi to sono miryoku (Nagano-shi: Isshōsha Shuppan, 2007).

138 specified territory.13 They comprise some of the most common surviving documents from the late medieval written period. Unfortunately, none of the confirmation documents from Nobutora’s 1522 gathering at Shōhōji survive. That in itself is not particularly surprising, as historical materials from the 1520s are extremely sparse, and there is a near total absence of Takeda-issued materials to retainers from 1518-

1522.14 But by examining confirmation documents from subsequent decades, we can understand the terms laid out both in Nobutora’s day, and that of his successors.

Confirmation often begins by noting that the terms of service shall proceed “as before”

(mae mae no gotoku 如前々). They frequently make explicit reference to service rendered to the previous Takeda family head. In other words, the first thing that these documents establish is that the terms of service (recognition of rights in exchange for military service) shall continue under the new lord. A 1543 confirmation letter issued to Takeda retainer Ozawa Miyabō 小沢宮坊 (unknown) by Shingen provides a typical example.15 Shingen issued this document to Miyabō upon the death of his father

Ozawa Bun’emon 小沢文衛門 (unknown) reauthorizing Ozawa’s retainer status, which had been established by Nobutora via letter (Nobutora yori hangata dasare sōrō

自信虎判形被出候).16 Thus, the Ozawa-Takeda relationship was authorized to

13 Or later, when few retainers held actual fiefs, rights to a specified stipend from the lord’s warehouse (okuradashi 御蔵出). This became typical in the seventeenth century, but there are examples from the later years of the Takeda administration in Kai. 14 Shibatsuji Shunroku, Kuroda Motoki, and Marushima Kazuhiro, eds., Sengoku ibun: Takeda-shi hen v. 1 (Tokyo: Tōkyōdō shuppan, 2002), 23–25. 15 Ibid., 71. 16 Not much is known about this Ozawa lineage other than that they hailed from Fujiyoshida and claimed the title Tanba no kami 丹波守 in the Edo period. They appear to have become Oyamada retainers in the 1550s, possibly as part of the Takeda granting the Oyamada increased control over Tsuru county around that time. Shibatsuji et al., eds., Takeda-shi kashindan jinmei jiten, 2015, 193–94.

139 continue without alteration. Documents from the next generation Takeda leader,

Katsuyori 武田勝頼 (1546-1582), followed the same formula. For instance, a confirmation letter to Osano Echigo no kami 小佐野越後守, a fairly high-ranking retainer of Tsuru County, came from Katsuyori immediately after he assumed family headship upon Shingen’s death in 1573.17 This letter assured Osano that there would be no change from the investitures granted by lord Hōsei’in 法性院殿 (Takeda

Shingen’s posthumous name) and that Osano’s status as a member of the military service group should continue as before.18

The Takeda incorporated people outside the military service group in largely similar fashion: confirmation of rights in exchange for specialized services, or depending upon the recipient’s occupation, goods. Some of Nobutora’s earliest sealed orders ever issued were confirmation documents to local temples in Kōfu. A 1517 document granted to Jishōji 慈照寺, and another the following year to Ichirenji 一連寺 endorsed the temples’ property claims (jiryō 寺領).19 Other confirmation letters of this era specify claims to particular resources. For instance, Nobutora gave a document to Kōgō-in 広厳院 in 1518 announcing a prohibition on the cutting of bamboo and lumber (chikumoku 竹木) from an unnamed forest “attached to” (tsuki

17 The Osano lineage were originally Oyamada vassals, and were granted hereditary status as head priests of Fuji Omuro Sengen Shrine 富士御室浅間神社 sometime during the age of Oyamada Nobuari Etchū no kami 小山田信有越中守 (1488-1541). They later received investiture directly from the Takeda. The recipient of this 1573 confirmation letter was Osano Yoshihide 小佐野能秀 (unknown). Ibid., 192. 18 Yamanashi kenshi shiryō-hen 4, 1112. 19 Jishōji is a Sōtō Zen temple in contemporary Kai City 甲斐市, to the northwest of Kōfu. Ichirenji is a prominent Jishū 時宗 institution in Kōfu’s Ōta-machi 太田町. Sengoku ibun: Takeda v. 1, 21–22.

140

付) the temple and its branch institutions.20 The order goes on to condemn violence and disorder over the issue of cutting, and that violators shall be punished. In this way, Kōgō’in gained a powerful ally in preventing precisely this kind of activity and bolstered their own claims to an important local resource. These arrangements were the foundation of Takeda territorial administration. They are the very first orders of business advanced by Nobutora, and continued to be so under his successors. Over the next several decades, confirmation letters became much longer and more specific, indicating a process of refinement over the years. From simple, one or two line proclamations, confirmation letters grew to substantial length and detail. Emblematic of this later phase is a 1575 document granted to Ochiai Village 落合之郷. It confirmed rights to harvest bamboo, cut trees, gather materials for rope, and take underbrush for use as fertilizer, among others.21 These terms are laid out item by item in a document roughly four times as long as the examples from 1517 and 1518.

While the Takeda granted confirmation of rights in exchange for service in their capacity as regional warlords, the same process took place at the county and local levels. Takeda retainers gained credibility and asserted authority over territory in much the same way as their superiors did. The first generation Oyamada Nobuari (see note 17), premier warrior landlord in Tsuru County began to do this in the 1530s. He granted exemptions from property tax, construction duty, and miscellaneous duties

(shoyaku 諸役) to a small-time retainer named Tarōuemon 太郎ゑもん (unknown) of

20 The prohibition would have been applied to “outsiders,” in other words, people who did not belong to the temple. Though the forest is unspecified, the language of this document indicates that it was one customarily claimed by the temple as under their purview. Ibid., 22. 21 Yamanashi kenshi shiryō-hen 4, 317–18.

141

Shishitomi 鹿留 in a confirmation letter of 1534.22 Oyamada, like their Takeda lords, offered tax-exempt status to those willing to provide military service. These arrangements were re-authorized by successive generations. The Osano lineage had received such re-authorization for several generations prior to the aforementioned

1573 confirmation letter of Takeda Katsuyori. Initially Oyamada retainers, a 1556 document from the third generation Oyamada Nobuari (Yasaburō 弥三郎, d. 1565) outlines various exemptions “as has been in past” (mae mae no gotoku).23 The document confirmed a property tax exemption on five residences (ie itsutsu 家五つ).

How far back such exemptions extended was not specified, but nonetheless these service terms had previously been established.24 They appear to follow the same convention of property tax exemption as the basic perk afforded to military retainers.

Some later Oyamada documents issued to retainers are more explicit in confirmation of rights for the type of service required. For example, a local warrior named Nakamura Yojūrō 中村与十郎 (unknown) of Kawaguchi village 河口郷 received a confirmation letter from his lord Oyamada Nobushige 小山田信茂 (1540-

1582) in 1568.25 In exchange for hereditary military service, Nakamura’s property, including a villa (yashiki 屋敷) in town and registered fields (myōden 名田), was hereby recognized as his by the Oyamada. Nobushige issued many similar documents to his retainers over the years. In 1573 Magoemon no jō 孫右衛門 of Ōhara 大原,

22 Ibid., 1082. 23 Ibid., 1106. 24 Shibatsuji et al., Takeda-shi kashindan, 192–93. 25 Yamanashi kenshi shiryō-hen 4, 1092.

142 where Oyamada headquarters lay, received confirmation of ownership (myōshu 名主) of a one kan fief, plus additional reward lands for meritorious military service, all to be held by descendants in perpetuity.26 Another small-time warrior, Saruya Iwami no kami 猿屋石見守 (unknown) of Kawaguchi gained exemption from various duties in

1574 for his diligent service to Nobushige.27 In that same confirmation letter

Nobushige admonished Saruya not to challenge or rebel against his superiors (ie, the

Oyamada). These documents clearly reveal very practical, quid-pro-quo arrangements that underlie late medieval vassalage. They also give a glimpse of the main issues and concerns to each party. In general, we can say that inferiors, at various socioeconomic levels, sought to have their claims endorsed by superiors who could function as influential allies or patrons. On the other side, superiors’ overriding concern was stability, or specifically, not rebelling against their authority, and clear pledges to uphold specified obligations, such as military service.

Recognition of local claims formed the foundation of late medieval warlord governments, and continued to be the basis of hierarchical sociopolitical relationships into the early modern period. Chapter two detailed how that relationship was established and how it might be used by subordinates to advance local interests in disputes. But how was this relationship supposed to function normally? In many ways, Takeda administration of Kai was a continuation of local practices that had been in place for a long time, subject to rather limited oversight by the lord’s officials. From the local perspective, daily life probably did not change much throughout the

26 Ibid., 1124. 27 Ibid., 1102.

143 sixteenth century, although the risk of violence gradually waned. Only two basic aspects of territorial administration under the Takeda set it apart from what came before. First, rights became centered upon physical occupation of a place. Second, obligations, including taxes and services, became centered upon a relationship with the Takeda. This amounted to a streamlining of a tangled web of rights and obligations that had prevailed from the late Heian period, and this proved to be more functional for a larger swath of the population.

The Takeda reorganized taxation in Kai from the mid to late sixteenth century.

While this certainly expanded Takeda economic clout, new tax structures primary historical significance is as a new mechanism for organizing rights and distributing surplus. As noted above, Takeda gave retainers exemptions in exchange for military service. However, most Takeda territory was occupied by people who were not part of the military service group. These individuals were classified according to their primary occupation, and in principle owed two kinds of tax to the Takeda: production tax, referred to by various names, and the property tax, called munabetsu 棟別. They also owed the Takeda “various duties” (shoyaku 諸役), which included corvee labor and a number of miscellaneous levies. The term munabetsu could in fact refer to either taxes or duties assessed to a household. Therefore, depending upon the context, a munabetsu exemption in Takeda documents might refer to the property tax specifically, duties, or some combination of the two. In these cases the terms of the exemptions and what they apply to are explicitly stated. Most times the munabetsu

144 refers to the basic property tax, which was correlated to the size and location of a residence.

Takeda property tax marks the first time that a levy was applied universally across a jurisdiction defined by territorial control. In other words, it was divorced from previous systems of authority centered upon the imperial or shogunal polities, and existed in the Takeda realm because the Takeda held sway there. Like other regional magnates, the Takeda only collected a production tax from lands that they “owned” directly.28 This was a continuation of the basic structure of the landlord-tenant relationship which had operated for centuries.29 The production tax therefore functioned more like rent, because it was paid to a landlord (whether on the local, county, or provincial scale) from his direct subordinates. That is to say, the occupants of his direct property holdings. The property tax, on the other hand, applied to everyone who did not have an exemption.30 It was not implemented until Takeda rule of Kai had been solidified over the course of several decades. Although applied piecemeal under Takeda Nobutora, the property tax was first implemented across the whole of Takeda territory in the 1540s with the completion of the so-called “property tax register” (munabetsu-chō 棟別張) in 1541.31 That was not coincidentally same year that Shingen assumed power by overthrowing his father. That it was one of

Shingen’s first acts as Takeda head suggest its importance and utility as a means to solidify Takeda rule. However, it is unrealistic to think that Shingen, who was in a

28 Sasamoto, Sengoku daimyō, 229. 29 Although who occupied the role of landlord and landlord agents did significantly change over the course of the medieval age. 30 Sasamoto, Sengoku daimyō, 232. 31 Ibid., 241.

145 precarious political position early in his reign, could dictate such a sweeping policy were its benefits not distributed.32 The property tax register might appear to be a new imposition created from whole cloth, but it was actually a redirection of older service obligations or of existing local practices. And, although non-retainers did not enjoy an

“automatic” property tax exemption, the terms of their obligation to the Takeda were in practice modulated by a variety of factors including occupation, location, and the kinds of goods and services rendered.

Taking a bird’s eye view of the Takeda tax system makes it appear quite systematic. But close examination of tax documents reveals that is was highly case- specific. How exactly Takeda lords, and smaller-scale lords, actually tapped into local surplus varied and was routinely adjusted. What this indicates is that these arrangements were significantly shaped by participants at both ends of the sociopolitical spectrum. Let us consider how tax exemptions were actually applied to various non-retainer groups.

The military service group was not the only one to be exempted from the production tax. A 1571 exemption offered to an organization of merchants based in

Kai offered remission of the production tax “as it is for the military service group”

(gun’yakushū no gotoku ni 如軍役衆).33 These merchants were engaged in long- distance shipping and likely provided the Takeda access to goods that were difficult to procure from within the province. Because of their ability to provide an essential

32 See chapter four for a discussion of the circumstances under which Takeda Harunobu assumed power. 33 Yamanashi kenshi shiryō-hen 4, 1134.

146 service, they earned a tax exemption essentially identical to that enjoyed by retainers.34 The Takeda offered a similar exemption to miners of Kurokawa gold mine

黒川金山 in central Kai (see map 3.1).35 Miners across the Takeda realm were employed in various military campaigns. These specialists proved so essential to the war effort that they were granted a kind of quasi-retainer status and classified as a special group, the kinzan-shū 金山衆.36

Takeda granted tax exemptions to individuals who were not hereditary retainers or specialized artisans as well. Some common recipients included religious institutions, artisans, and even ordinary villages. For instance, an undated document of Takeda Nobutora, likely issued in the 1530s, granted a full property tax exemption to Katsunuma village 勝沼郷.37 In this case, the village must have provided a special service, most likely portage or some other kind of logistical support for a military campaign. This shows that villages too, like military specialists, were sometimes co- opted for service with special privileges. Early exemption documents are short and tend to offer limited detail. But by the 1580s, when the tax system was fully

34 The only difference was that it was not hereditary and intended to be applied in perpetuity. These arrangements were often in effect for a specified number of years, although that was not the case with the example given here. 35 Yamanashi kenshi shiryō-hen 4, 520. 36There is quite a bit of debate amongst medieval scholars regarding the nature of Kai miners and their relationship to the Takeda. Some consider them to be direct retainers of the Takeda just like warriors. Others argue that miners were largely independent and classified as artisans from the Takeda perspective. My position is that although they were explicitly not considered military retainers, based upon the wording of various documents, in effect they held quasi-retainer status because they participated in campaigns and enjoyed the same tax exemptions as warriors. For a summary of this debate, see Nishikawa, Chūsei kōki no kaihatsu, 142–49. 37 Yamanashi kenshi shiryō-hen 4, 1105. Katsunuma is now part of Kōshū City.

147

Kurokawa Gold Mine

Map 3.1: Gold Mines of Kai Province38

developed, detailed registers such as the one issued to Iguchi village 井口郷 emerged.39 This lists first the households in Iguchi that hold exemptions. Judging from the names, members of this group are lesser military retainers, beginning with the highest status individual, one Imafuku Izumi no kami 今福和泉守 (Masatsune 昌常, unknown). Next are the details of the taxes to be paid, divided into three categories: main houses, new houses, and one unoccupied villa. This document was addressed to

Iguchi village itself, meaning the village council. That indicates that residents

38 Ibid., 140. Note the comparatively small number of mines in the northwest of the province, where the Takeda were most influential. The vast majority of medieval gold mines in Kai were in Anayama territory in the south. 39 Ibid., 331. Iguchi is now part of Chūō City.

148 themselves were collectively in charge of the implementation and collection of this tax. Exemption documents from this era therefore explicitly reinforced local authority in matters such as tax collection and the allocation of duties.40 Furthermore, as was the case for taxes, the actual imposition of duties was subject to case-specific exemptions, reductions, or other modifications. These were based not on property but upon the type of service rendered to superiors. In order to examine how duty exemptions were actually implemented, we must consider the discourse of obligations developed in Kai under Takeda rule.

B. Reorganization of Service Obligations

New kinds of taxes created a new system of obligations and responsibilities. In Kai by the mid sixteenth century, in principle all residents of the Takeda domain owed some kind of service obligation to the Takeda. This was quite a significant departure discursively; however several points must be emphasized regarding the implementation of this new sociopolitical system. Takeda policies did not create new occupational groups. High ranking warrior households such as the Takeda and

Oyamada did not bring into existence new ranks of local warriors by decree. They rather recruited specialists to their cause through the basic confirmation for service arrangement described above. This same process took place with other occupational groups. And, although non-military retainers did not ostensibly enjoy tax exempt status, recruitment of service from non-military occupations also followed the basic

40 Sasamoto, Sengoku daimyō, 233.

149 structure of confirmation for service. While this process did stimulate economic growth and redirected surplus in various ways, it did not create new sectors in the late medieval Kai economy.41 Elites turned to existing sources for the things they needed and found ways to establish exchange relationships. This was most clearly evident with artisanal organizations whose products or services were essential for elite military campaigns. However, as we will see, that structure was applied in practice to all occupational groups, regardless of status, to varying degrees. This had the effect of rendering elite authority both less invasive, from the local perspective, and expedient towards certain goals.

In addition to local tradespeople, several large, inter-regionally active guilds (za

座) operated in late medieval Kai.42 In both cases, property tax or duty exemptions were granted in exchange for raw materials or labor.43 Each occupational group rendered service to the Takeda or Takeda-backed retainers, and this confirmed that these artisans had the rights to ply their trade. In this way, the relationship was structurally similar to the master-retainer relationship of warriors, and the lord- resident relationship in local communities.

One of the most important occupational groups for the Takeda was lumberjacks

(yamazukuri 山作). This trade was essential for the military effort and everything else

41 Sasamoto, Sengoku daimyō, 270. This point is contrasted with the qualitative economic change of the seventeenth century, which included new products, new trades, new ways of specialization, long-distance trade, and consumerism. Overall, the early modern economy can be characterized as commercial, whereas the late medieval remained based upon the local and regional markets. 42 The difference is most clearly evident in documents addressed to artisans with surnames (heads of large organizations) and those without. Sasamoto, Sengoku daimyō, 275. 43 For a list of Takeda-issued documents to artisan groups and the kinds of exemptions offered, see Ibid., 270–71.

150 that depended upon a supply of building lumber. In the late medieval economy, harvesting timber, shipping it, and selling it at market had yet to be divided into separate occupations. Specialists who cut large timbers also shipped and wholesaled them. Consider the “Western Sea Group,” an organization of lumberjacks based in central Koma County, in the forests directly west of Lake Motosu. They cut timbers and shipped them through the corridor running along Mt. Fuji and medieval Yoshida.

As early as 1540, this group of lumberjacks enjoyed an exemption from a checkpoint toll along that route from Takeda Nobutora.44 Thirteen years later, Takeda Shingen laid out exemption terms to eight leaders of this guild in a 1553 order.45 It specified that they were to continue supplying timber to the outpost (ban 番) at Motosu, and that in return they shall receive toll exemptions for complete round trip journeys to and from Fuji (ie Yoshida village). This arrangement was specified to be valid for the coming ten years. This is a typical example of an artisan organization involved in a particular business enterprise that was co-opted to assist the Takeda with a specialized raw material. In this case, the Western Sea Group were the main suppliers of lumber to Yoshida. They were granted perks in exchange for supplying the outpost at Motosu, a point along their normal route, with a portion of their product. It is easy to see the mutual economic benefit both parties enjoyed from such an arrangement.

But on a more fundamental level, this supplier-exemption relationship also established that the Western Sea Group had the rights to take timber from the forests at Motosu and move them into Yoshida. Takeda authority thus endorsed the

44 Yamanashi kenshi shiryō-hen 4, 1116. 45 Ibid.

151 lumberjack guild’s claims to resources. Southwest of Motosu, the county-level warrior house Anayama established the same kind of relationship with a lumberjack guild headed by the Sano 佐野 lineage.46 Local-level elites did the same in their own areas.

Trades that supplied skilled labor rather than raw materials usually received exemptions from corvee labor/construction duty (fushinyaku 普請役). Notable examples from sixteenth century Kai pertain to carpenters (daiku/saiku 大工・細工), another essential occupational group. During the building boom at Kōfu in the mid sixteenth century, the Takeda granted numerous exemptions to builders. For instance, in 1568 the members of the Okura 小倉 guild received a reduction in construction duty equivalent to the value of a one ken dwelling.47 Their services must have grown increasingly valuable to the Takeda, for in 1576 they were granted a total exemption from construction duty in perpetuity by Takeda Katsuyori.48 Other Kōfu carpenters, such as the Kagami 加賀美, received similar tax reductions or exemptions at this time.49 The Kurosawa 黒沢 carpenters of Kawaguchi, north of Yoshida, likewise received exemptions from the Oyamada. Here too, it seems their services become more and more essential over the years. In 1568 Oyamada Nobushige issued property tax exemptions for two houses.50 By 1577 the Kurosawa enjoyed a total property tax exemption, to be held in perpetuity.51 In all of the above cases, influential warrior

46 Nishikawa, “Sengoku-ki Kai no kuni ni okeru zaimoku no chōsetsu to yamazukuri,” 5. 47 Yamanashi kenshi shiryō-hen 4, 316. 48 Ibid., 315. 49 Ibid., 321. 50 Ibid., 1094. 51 Ibid.

152 lineages recruited an established artisan organization into their service by offering economic incentives.

Probably the most famous regional artisan group in Kai was gold miners. Scholars have long debated the nature of gold mining in the province, and disagreed over the relationship between miners and territorial lords.52 A longstanding historical theory posits that the Takeda were able to seize control of gold mines in Kai quite early in their reign. Ownership of these mines directly funded the Takeda military and served as the key financial resource which allowed the Takeda to establish superiority across

Kai. This assessment is largely unsubstantiated in the historical record, and now questioned by many experts. For starters, documentation proving Takeda control of the several dozen gold mines in Kai does not exist (see map 3.1). Instead, most of the documentary evidence indicates that gold miners were taxed in much the same way as other artisanal groups described above. Most Kai gold mines rested within

Anayama territory in Koma County, and it was the Anayama, not the Takeda, who appear to have had the most extensive relationship to miners in the sixteenth century.53 However, even the Anayama only “owned” mines in the sense that they were entitled to a percentage of miners’ yields at sites within their territory, in exchange for various exemptions. That was the arrangement the Anayama had with miners of the Tsuzura 黒桂 gold mine in western Koma.54 Political elites do not appear to have been directly involved in mine operation until Tokugawa control of Kai

52 The nuances of this complex historiographical debate are beyond the scope of this chapter. For a summary, see Sasamoto, Sengoku daimyō, 304–22. 53 Nishikawa, Chūsei kōki no kaihatsu, 140. 54 Sasamoto, Sengoku daimyō, 284.

153 in the early seventeenth century.55 At that time, circulation of the famous Kōshū-kin

甲州金 increased dramatically. Sixteenth century examples of the coin are extremely limited, and were likely used as gifts from the Takeda to religious institutions or in some cases retainers. They were not regular currency.

The best documented gold mine in Kai was the Kurokawa mine in northwest

Kuninaka. Even still, many uncertainties remain. The basic relationship established with Kurokawa miners by the Takeda was essentially the same as that with lumberjacks and carpenters. Gold miners too both procured their raw materials and shipped them, and as a result the Takeda offered toll exemptions in exchange for a percentage of the mineral yield. For instance, in 1571 Tanabe Shirōzaemon no jō 田辺

四郎左衛門尉 of the Kurokawa miners guild received an exemption from tax on one horse load (uma ippiki 馬一匹) of yield per month.56 This same order granted miners a full property tax exemption, a production tax exemption, and an exemption from construction duty. Basically, the Kurokawa miners had all of the privileges of military retainers, leading some, such as Sakurai and Shibatsuji, to conclude that they were.57

But several points suggest an alternate conclusion.

Kurokawa’s exemptions show that the miners owned productive land. This suggests that the Kurokawa guild was not solely comprised of miners, but a number of other occupational groups who supported the mining operation. This too, sounds

55 Sasamoto has suggested that control of specie is therefore in fact one of the distinguishing characteristics between Sengoku daimyō and Kinsei daimyō. Ibid., 314. 56 Yamanashi kenshi shiryō-hen 4, 426; Nishikawa, Chūsei kōki no kaihatsu, 144. 57 Nishikawa, Chūsei kōki no kaihatsu, 145.

154 largely similar to warrior landed property. However, the exemption document presented to Tanabe in 1571 explicitly noted that the Kurokawa guild shall be granted a production tax exemption like that of the military service group (gun’yakushū no gotoku 如軍役衆).58 It turns out it from the Takeda perspective, miner’s most valuable skill had little to do with precious metals themselves. It was prospectors’ excavation expertise that was so useful to the war effort. According to the 1571 document, the miners participated in the siege of Fukazawa castle 深沢之城 in northern Sagami

Province.59 Though not specified in the 1571 exemption, Kōhakusai’s journal reveals the miners’ role in the conflict. They were employed in numerous sieges for excavation of defensive works, fortifications, siege tunnels, well-digging, and the re- routing of waterways. Over Kōhakusai’s roughly two-decade account of military service, he mentioned cutting off a castle’s water supply (mizu no te 水ノ手) on several occasions. In 1546 Takeda forces took the water from Maeyama castle 前山城 in Saku County, Shinano, and employed the same strategy the following year at Shiga castle 志賀ノ城.60 Denying castle defenders water was unsurprisingly a common tactic in late medieval warfare. It was miners, at least for Takeda forces, who actually performed the labor of doing so. Kōhakusai’s journal also routinely describes the digging of defensive earthworks during sieges and groundbreaking ceremonies to

58 Yamanashi kenshi shiryō-hen 4, 520. 59 Fukazawa castle was originally built by the Imagawa, then taken by the Takeda in 1568. The Hōjō captured it from Shingen in 1570, who then launched a successful counter-siege in the 1571 campaign referenced here. 60 Kōyō nikki, 89-90.

155 initiate new construction. Most Kai specialists conclude that miners were employed as excavators for these projects.61

Elites’ role in promoting gold mining in Kai has likely been overestimated, and precious metals were probably less important to the Takeda overall economic management of their territory than has been argued. Consider an undated document thought to be from the mid Tenbun era (1532-1554) issued to the head of Omura

Sengen shrine in Yoshida. This institution collected fees from pilgrims to Mt. Fuji, hosted markets, and sold religious items such as omamori 御守り, a type of talisman.

All of this commercial activity meant that a substantial amount of cash flowed in and out of the temple each year. Debased currency, so called “bad coins” (bitasen 悪銭) was a consistent problem for late medieval commerce due to the shortage of coins.

Myōhōji-ki indicates that buying and selling in Yoshida was repeatedly hindered by the prevalence of bad coins. This was primarily a concern for religious institutions, the main entities engaged in market activity in the area. Omuro must have had to deal with this problem on a regular basis. But bad coins could still be valuable raw material, indicated by the Takeda offer to Omuro that the shrine send bad coins to

Kōfu to be melted down for musket balls, in exchange for gold or property tax exemptions.62 Clearly the Takeda had access to gold. Acquisition of base metals for weapons production appears to have been a more pressing concern than accumulation of bullion.

61 Sasamoto, Sengoku daimyō, 330. 62 Yamanashi kenshi shiryō-hen 4, 1114.

156

For all of the scholarly attention devoted to warlord-promoted development in the sixteenth century, the examples from Kai indicate that the Takeda and other political elites tapped into existing economic networks. Their influence grew not because they stepped into chaotic situations in the countryside and took over direct control of local surplus. Instead, their authority was based upon reinforcing local practice and maintaining stability. Takeda implementation of the property tax, for example, was always adjusted in ways that were largely in line with local practice, while still allowing for superiors to receive desired economic resources. There are many examples of adjustments, which appear one-sided because surviving sources were written by the Takeda government and mostly only stated its interests. But these arrangements were in fact worked out as the result of significant input from sociopolitical inferiors. Illustrative of this process is the production tax exemption given to Mitake 御嶽 in 1564 by Takeda Shingen.63 Located on the outskirts of Kōfu, this area was considered “backwoods” (sanchū 山中) and would not have normally been subjected to duties associated with an established village.64 Mitake’s terms of incorporation under Takeda rule reflected that fact: residents were exempted from construction duty. The Takeda did not “normally” grant settlements such exemptions, although there were ways for villages to receive them. If a group had a reasonable justification for tax exemption, it seems the Takeda generally took the path of appeasement. For instance, in 1577 nine adjacent villages in Tsuru County received exemptions from the property tax because of their collective participation in

63 Ibid., 322. 64 Known by various names in the medieval era, this came to be referred to as mura-yaku 村役 in the Edo period and it entailed contributions for the upkeep and administration of communal property.

157 unspecified construction duty.65 These are privileges that locals petitioned for and won. It is not wonder that the Takeda were able to incorporate new territory relatively smoothly.

Most of the obligations required of subordinates by the Takeda and other territorial lords were not new, either. Religious institutions gained confirmation and other perks simply for acknowledging Takeda supremacy and continuing to provide services that they were already engaged in. Takeda Shingen is famous for new law codes (hattō 法度) issued to temples and shrines in Kai. But these were only new in the sense that they were coming from a regional warlord, a warrior, and not from elite central religious authorities. They did not fundamentally alter the inner workings of these institutions.66 For example, a 1561 law code issued to Kōfu Hachiman Shrine 八

幡神社 contained detailed rules for day-to-day activities of shrine personnel.67 These rules would have been crafted by shrine leadership themselves, and endorsed by

Takeda authority. This meant that shrine leaders could bring Takeda influence to bear should their own authority be challenged or if disciplinary action was necessary.

Temples and shrines also received exemptions for services in much the same way as local communities. In 1575, Entakuji 塩沢寺 in Kōfu received a typical investiture of this kind from the Takeda. 68 The document first confirmed the temple’s property claims, then granted a property tax exemption of three ken per house for the

65 Yamanashi kenshi shiryō-hen 4, 1117. 66 The establishment of stability in Kai did, however, eventually render the military forces of religious institutions obsolete, as it did for local communities. This development did result in a firm subordination of religious political authority to that of the warlord class in the long term. 67 Yamanashi kenshi shiryō-hen 4, 298–303. 68 Ibid., 284.

158 dwellings of temple personnel. Entakuji’s basic needs were met and it was business as usual, in exchange for acknowledgement of Takeda authority and a limited payment.

The Takeda have been considered progenitors of a widespread construction boom in Kai from the 1550s onwards which resulted in long-term improvements. But here too, their significance was primarily as a facilitator for the successful execution of local infrastructure management. Takeda warlords could coordinate local projects to achieve larger, regional goals, and they helped resolve disagreements over the implementation of large scale building. As in local resource disputes, the Takeda’s most important role was as a stabilizing force capable of keeping local conflicts in check. This created an incentive for locals to see building projects successfully completed. It also meant that everyone, even the Takeda, had obligations towards the maintenance of order. All were subject to the new stability. We turn now to the implementation of several of the most famous projects in sixteenth century Kai

Province.

C. Construction Duty in Late Medieval Kai

From the 1540s on, because of Takeda success in managing loyalties within Kai, military campaigns were almost exclusively external. That meant that for the most part, residents of the province no longer had to deal with incursions by hostile armies.

This contributed greater incentive to invest in expensive, time-consuming improvements to local infrastructure. Effective dispute resolution meant that communities could rely upon agreements worked out with neighbors. Better

159 coordination led to the integration of projects across several adjacent areas. The

Takeda wished to encourage such cooperation and authorized larger, more ambitious projects over the course of the sixteenth century. From the perspective of political leaders, mobilization of locals for construction duty looks like an application of coercive power. There certainly was a coercive element, but it came from both sides of the sociopolitical spectrum. Infrastructure projects such as river dredging, dam building, or road maintenance tangibly benefitted local communities as well.

Participation also conferred upon locals additional claims to resources and territory.

Communities essentially behaved as the “owners” of territory that they were charged with maintaining. This was precisely the logic employed by Yoshida village in their

1635 suit against Kurechi and Onuma. In that instance, Yoshida argued that the areas in which they and Kurechi were responsible for road construction amounted to a tacit acknowledgement of the border between the two villages.69 Crucially, locals could depend upon the participation of other parties in Takeda-backed construction projects. Benefits were therefore on three interconnected fronts: tangible improvements to local and interlocal infrastructure, additional backing of local resource claims, and cooperation with outside groups on specified terms.

The Takeda were clearly involved in certain kinds of construction projects during the sixteenth century. They provided the major impetus for creation of the network of outposts and garrisons that made up the post-horse system, which was initiated in the

1520s. Takeda Shingen famously moved Zenkōji from Nagano City in Shinano to his

69 This was one of several points that Yoshida villagers made in order to assert their claims to a piece of land called Shimo no mizu. See chapter 2 for detailed analysis of this event.

160 capital at Kōfu in 1558. That was undoubtedly a major undertaking and would not have happened were it not for the ambitions of the Takeda warlord. Likewise, Takeda

Katsuyori’s construction of Shinpu Castle 新府城 at Nirasaki 韮崎 in 1582 was based upon geopolitical and military strategy, not exploitation of resources per se. These could all be said to be Takeda projects. But infrastructure building is something else.

Infrastructure projects made various useful improvements to local environments for the people living there. When we examine these undertakings, the Takeda role appears far less active. Evidence linking the Takeda to the two most famous projects of late sixteenth century Kai is surprisingly slim, suggesting their primary role was as facilitator, and, as in resource disputes, guarantor of local claims.

If Kai itself is inextricably linked to Takeda Shingen, it is safe to say that Shingen himself is inextricably linked to the two most famous construction projects of the sixteenth century: the bō-michi 棒道 and Shingen’s Dike (Shingen tsutsumi 信玄堤).

These projects are considered Shingen’s most lasting contributions, nearly as well- known as Kawanakajima and the rivalry with Uesugi Kenshin 上杉謙信 (1530-1578).

The bō-michi and Shingen’s Dike are widely considered to have been completed in

Shingen’s era, circa 1560s, but scholars have debated the genesis of these projects for decades.70 For all the credit heaped upon the Takeda lord, it is highly unlikely that

Shingen initiated or directed these projects. Most contemporary scholars conclude that the projects were made possible by an environment of greater stability in mid

70 For a summary, see Sasamoto, Sengoku daimyō, chap. 1.

161 sixteenth century Kai. They were expansions of much older local environmental management efforts.

The lack of clear evidence pointing to Takeda direction of the road and dike points to Takeda as facilitator rather than progenitor.71 Supposedly, the bō-michi was constructed in the late 1560s on Shingen’s orders, to facilitate swift movement of troops between Kai and Shinano.72 Some have even speculated that Shingen anticipated the protracted showdown with Uesugi Kenshin, and thus invested in this project. The road was so named either because it was a relatively straight shot between the two provinces, or because it had fairly small elevation changes. Whatever the case, there is virtually no reliable historical evidence for the bō-michi. One of the only documents pertaining to it comes from the collection of the Takamizawa lineage

(Takamizawa monjo 高見沢文書) and is for various reasons of dubious authenticity.73

Nonetheless, the road continues to exist today, and archaeological analysis suggests construction was indeed at some point in the sixteenth century. Had the road been completed later, chances are that more documents would have survived pertaining to its construction. As it is, the only firm conclusions to be drawn are that it was built at some point during the Takeda rule of Kai, and that specifics cannot be historically verified. Even considering the sparse surviving evidence from the time period, we

71 When we consider the body of Takeda documents as a whole, it become very unlikely that evidence for Takeda participation in these two projects would be so slim had their role in them been more active. 72 Ibid., 48–57. 73 Yamanashi kenshi shiryō-hen 4, 807; For a discussion of the problems with this source, see Sasamoto, Sengoku daimyō, 58–61.

162 would expect to see clearer indications of the project’s genesis if it was initiated by

Shingen. Who built the road and how simply cannot be determined.

The historical details of Shigen’s dike are frustratingly ambiguous as well.

Recognized as one of the greatest engineering achievements of the sixteenth century, the dike protects Kōfu from serious flooding.74 It is located at the confluence of three large rivers, the Kamanashi, Shio 塩, and Midai/Midei 御勅使, in a particularly flood- prone section of the Kōfu basin (see map 3.2). The Shio River flows into the

Kamanashi at the northwest corner of Kōfu, which is then joined by the Midai only a few hundred yards downstream at a large bend. When these rivers are high, there is a lot of water converging and changing direction at the same time, leading to frequent floods. When this area does flood, it has the potential to reach deep into the city of

Kōfu, as it did in Keichō 慶長 13 (1608), when nearly half the city was underwater for some time.75 Like the bō-michi, there is no definitive historical evidence regarding

Shingen’s Dike. Links to the Takeda ruler himself come from oral tradition and the nineteenth century ethno-history Kai kokushi.76 Tangential documentation and archaeological analysis that cannot be precisely dated has led most scholars to conclude that the dike was built at some point between 1560 and 1582.77 That would indeed place it within Shingen’s era, but again the extent and exact nature of Takeda involvement in the project remains speculation.

74 For technical details of the dike itself, see Yamanashi kenshi tsūshi-hen 2: chūsei, 404–19. 75 Ibid., 408. 76 Ibid. 77 Ibid., 405.

163

Shio River

Kamanashi River

Shingen’s Dike Midai River

Kōfu

Map 3.2: Shingen’s Dike and Environs78

Our most tantalizing bit of historical evidence pertaining to Shingen’s Dike are

Takeda-issued construction duty orders from the Eiroku era (1558-1569).79 These orders are undated, but based upon their appearance and the seal, they are typically considered to be from Eiroku 6 (1563).80 One of these orders was issued to fourteen different villages, plus a group of villages known as the Hosaka association (Hosaka sōgō 保坂惣郷), all of which are located on the east bank of the Kamanashi directly in the flood plain (see map 3.3). All recipients were required to provide laborers

(ninzoku 人足) for flood prevention (tō mizu wo doku beshi 可退当水). The second

78 Google Earth. 79 Yamanashi kenshi shiryō-hen 4, 950–51. 80 Nishikawa, Chūsei kōki no kaihatsu, 246.

164 document was issued to nine villages further upstream the Kamanashi, with the same content.81

While we cannot say for sure that this was the order which initiated construction on Shingen’s Dike, these documents do reveal several interesting points about late medieval infrastructure building in Kai. They show both longstanding continuities in local administration and a growing capacity for interlocal cooperation, backed by

Takeda guarantees. The fourteen settlements of the first document had belonged to

Shima Estate 志摩荘 from at least the late twelfth century.82 Likewise, the Hosaka association consisted of fifteen villages which had comprised Hosaka Estate 穂坂荘

(see map 3.3).83 In this case villages were mobilized for construction not merely because of their location on the floodplain. When it came time to actually organize labor for the project, the Takeda relied upon existing local networks.

Even though the Takeda relied upon old Estate relationships to mobilize local labor for riparian works, implementing such projects was impossible prior to the developments of the late sixteenth century. One key ingredient was absent before then: locals’ ability to ensure the participation of all parties. By the second half of the

81 Yamanashi kenshi tsūshi-hen 2, 952–53. 82 Shima Estate’s original proprietorship appears to have been held by Matsuo Grand Shrine 松尾大社 in Kyoto, although it may have been associated with the aristocrat Fujiwara no Tadazane 藤原忠実 (1078-1162). It was held by Ichi’on-in 一音院 of the elite Kujō 九条 family in the later Kamakura era, and appears to have disentigrated by the late fourteenth century. Nishikawa, Chūsei kōki no kaihatsu, 286–87. 83 The villages of Hosaka Estate are mentioned in Kai kokushi, but little is known about the Estate’s history. Ibid., 250–51.

165

Group B Villages (above the floodplain)

Shingen’s Dike

Group A Villages (in the floodplain)

Map 3.3: Villages Involved in Ryūō River Works84

84 Ibid., 246.

166 sixteenth century, the Takeda could fulfill this role. The second Eiroku 3 construction duty order reveals how that took place. This document was in fact addressed to nine of the fifteen total Hosaka association villages. If these villages had been ordered to do construction duty via the first document, why would they be told to do so again? And why would the remaining six villages of Hosaka not be addressed in this follow up order? We can construct a theory by considering local interest. The six “unaddressed” villages (group A, map 3.3) were located directly on the banks of the Kamanashi.

Clearly in the flood plain, residents here stood to benefit directly from the construction of a dike. On the contrary, the nine villages singled out in the second

Eiroku document (group B, map 3.3) were located further upstream and at higher elevation in the hills east of the river. Residents here were likely much less concerned with flooding of the Kamanashi. It probably would have been very difficult to persuade these villages to contribute to a large undertaking such as a dike without an influential authority such as the Takeda. The nine villages of document two likely asserted that they should not be included in the construction project because of their location, but the Takeda responded with explicit orders to do so.85 In other words, the

Takeda made use of an old local network for a new purpose. Through a combination of coercion and incentive, the Takeda could make participation worthwhile.

The three major advantages listed above made participation in Takeda-backed projects worthwhile for most residents. Some saw further benefits, which followed the oft-used pattern of tax exemptions in exchange for service. An unaddressed

85 Ibid., 251.

167 document of Eiroku 3 (1568, the same year as the two mobilization orders discussed above) granted a full property tax exemption for participation in the “Ryūō riparian construction” (Ryūō no kawayoke 龍王之川除).86 This exemption was likely given to the group of retainers addressed in a Tokugawa-issued document of 1582.87 That order confirmed exemptions held by the “Ryūō riverbank outpost group” (ryūō kawara juku-shū 龍王河原宿衆) for participation in riparian works (kawayoke no hōkō 川除之奉公) should continue as they had “in Shingen’s time” (Shingen-dai no gotoku 如信玄代). We cannot know for sure if this exemption pertained to construction of the famous dike. But we do know that the dike was built, and that retainers usually received exemptions for particular kinds of service. It is therefore quite possible.

Implementation of old administrative structures towards new goals can be detected by examining the expansion of the Kamijō sluice 上条堰 under Takeda rule.

Although the sluice had existed since the Kamakura period, Sengoku-era modifications greatly expanded the structure.88 Typical of infrastructure building in

Kai during this era, the Takeda facilitated local efforts by coordinating construction and redirecting labor towards a common purpose. Located in what is now Kai City, the Kamijō sluice is a huge water conduit that supplied hundreds of outlets across medieval Kōfu. Takeda mobilized labor for a major rebuilding and expansion project in Genki 元気 3 (1572). Six villages received orders to repair a number of channels at

86 Yamanashi kenshi shiryō-hen 4, 950. Ryūō is the name of the area in Kōfu directly east of Shingen’s Dike. 87 Ibid., 954. 88 Nishikawa, Chūsei kōki no kaihatsu, 287.

168

Kamijō.89 But the details and implementation of this project had been decided based upon “consultation between the villages” (gō-chū dangō itashi 郷中致談合). Kamijō expansion was not merely a scheme cooked up by Takeda officials to improve water supply in northwest Kōfu. It was a collaborative effort between several local communities and authorities. The six villages involved in the Kamijō project had all been part of the aforementioned Shima Estate. These particular settlements continued to share a tutelary shrine in the sixteenth century, evidence of their continuing collective identity.

Completion of a big, expensive project such as Kamijo required cooperation, trust, and recognition of other parties’ rights and obligations – precisely the things that were so difficult to ensure during the Civil War era. Similar to their role as dispute mediators, the Takeda became the crucial third party in projects like Kamijō with the ability to serve as mediator, prevent violence, and ensure that participants both met their obligations and received their due. And indeed, a host of issues had to be sorted out when there were changes to the local environment. Dozens of Takeda documents of the late 1570s and 80s concern usage rights to areas opened up by new construction. Some of these locations are fairly straightforward, others much more specific. For instance, a 1575 document given to Kawara-juku village 河原宿之郷 at

Ryūō, granted rights to growth along the top of the dike, which villagers used for bamboo, fuel wood, straw, and rope making.90 Each major building like Shingen’s Dike or Kamijō created new opportunities and a new set of claims that had to be negotiated

89 Yamanashi kenshi shiryō-hen 4, 282. 90 Note that this was issued to the village (gō) and not the retainers (shū) at Ryūō. Ibid., 951.

169 somehow. These large scale projects would likely not have been possible if disputes could not be regularly resolved without serious disruption. Once those mechanisms were in place, it opened up possibilities for more extensive cooperation. These trends contributed to the realization of complimentary interest both vertically along the sociopolitical hierarchy, and horizontally as communities could rely on each-other to cooperate on specified issues.

Clearly there was new work being done in late sixteenth century Kai which resulted in several large scale building projects. However, it was not completely the result of new obligations placed upon subjects by rulers. As argued above, the Takeda found ways to tap into local manpower and direct it towards larger collective goals.

Management of local infrastructure evolved over the course of the medieval period and was intimately related to methods of conflict resolution. As historian Sano

Shizuyo has pointed out, early medieval building projects in the Kinai region tended to be for the village itself or tutelary shrine which might be shared with a handful of neighboring communities. The late sixteenth century is notable for the dramatic increase in interlocal projects.91 That was certainly the case in Kai, made possible because of interrelated increases in stability, assurances of claims, and guarantees of cooperation. Though facilitated by the Takeda, realization of these three trends was not only a function of Takeda power. Construction projects grew in scale and frequency along with Takeda influence in Kai. However, so too did the compromises and various mutually beneficial arrangements by which these projects were

91 Sano, Chū-kinsei no sonraku to suihen no kankyōshi, 211.

170 implemented. The key variable was not so much Takeda’s coercive force, but reliability – bolstered by the formation of new precedents.

Where the Takeda had the most direct impact upon regional development was by using their military retainers as a kind of crack construction crew. The military morphed over the course of the sixteenth century from a loosely coordinated band of warriors to an organized corps. During the first half of the sixteenth century their primary activities were property destruction (economic warfare), fighting, and building. It the second half of the century that order was reversed, with construction now comprising the majority of retainers’ service obligations. For Takeda bannerman

Komai Kōhakusai, who chronicled his days as a retainer from the 1490s to 1550s,

“going on campaign” (shutsujin 出陣) entailed roughly equal parts burning, battling, and building.

Kōhakusai and his comrades did a considerable amount of pillaging in their service to the Takeda. Arson was a favored tactic, employed frequently in conquest and siege. From circa 1520 to 1556, Kōhakusai recorded no less than a dozen instances in which Kai forces employed various scorched-earth tactics during the invasion of Shinano.92 These activities were not confined to what we would now consider to be “military targets.” More often than not, it was the countryside and ordinary settlements which suffered. Kōhakusai’s account of the war against the

Ogasawara, a county-level warrior lineage in central Shinano, provides a vivid example. In 1545-46, the Takeda army fought a major battle against an Ogasawara

92 Kōyō nikki.

171 retainer named Kasahara Kiyoshige 笠原清繁 (1515-1547) in Saku County 佐久郡.

Kiyoshige lost his stronghold Shiga Castle 滋賀城 and was killed in the fighting.93

Afterwards, Takeda warriors rounded up local residents and hauled them back to Kai as prisoners, where they were ransomed back to family members in Shinano for hefty sums of cash.94 Four years later in 1550, Takeda warriors had encircled the main

Ogasawara stronghold at Fukashi 深志.95 Kōhakusai reports that from their camp at

Yugawa 湯川 (modern Chino City 茅野市) to the southwest, they set out towards

Fuchū 府中 (Matsumoto) and burned fields of the surrounding villages along the way.96

Scorched earth tactics appear to have been widespread in Civil War era Japan.

Takeda Katsuyori described the status of his 1574 campaign against Tokugawa Ieyasu in a letter to retainers, matter-of-factly referring to the pillage and plunder involved.

Regarding a series of battles in Hamamatsu 浜松 of Suruga Province, Katsuyori stated that “all over the countryside, the homes of the people have been burned, with not one remaining. The rice plants, too, have been cut down and discarded, and [we’ve] thus achieved our main purpose.”97

Sixteenth century campaigns were thus as much about property destruction as killing enemy warriors. At the same time, however, the Takeda military was consistently building and rebuilding in conquered territory. Kōhakusai’s journal

93 Ibid., 95. 94 Myōhōji-ki, 35. 95 Later known as Matsumoto Castle 松本城. 96 Kōyō nikki, 95. 97 Yamanashi kenshi shiryō-hen 4, 1077.

172 contains many references to kuwa-date 鍬立, a kind of ground breaking ritual performed when a new structure was to be built. These ceremonies are frequently referenced in the wake of battles or sieges, suggesting that one of the first things the

Takeda army did after gaining territory was building the desired infrastructure. For example, after Kai forces took Takatō Castle 高遠城, in southeastern Shinano (modern

Ina City 伊那市), there was a groundbreaking for major modification and expansion of the fortification in the spring of 1547.98 The same thing happened in 1550 after

Ogasawara Nagatoki 小笠原長時 (1514-1583) lost Fukashi Castle to Shingen.99 In addition to military fortifications, the diary is littered with references to the building of mansions, villas, and religious institutions, all as part of Kōhakusai’s service to the

Takeda. He was even placed in charge of a river maintenance project at Obu 飯富 in

1546.100 Basically, warrior bands were the Takeda’s build crew as well as their muscle.

Outside of campaign season, retainers were still expected to participate in construction duty, and were only exempted from it in special circumstances. Such an exemption, which amounted to a reduction in service obligations, could only be earned through additional service. For instance, Shingen handed out exemptions in

1571 to about a dozen retainers who had served as his attendants to his wife, Lady

Sanjō 三条殿 (1521-1570).101 The Odagiri 小田切 lineage earned construction

98 Kōyō nikki, 90. 99 Ibid., 95. 100 Ibid., 89. Obu is located in southern Koma County, and the work probably was done along the Fuji River 富士川. 101 Yamanashi kenshi shiryō-hen 4, 279, 281.

173 exemption too, for their service to the Takeda lord’s immediate family.102 But under normal circumstances, retainers were expected to contribute labor in much the same way as non-military residents. One such order comes from Mizushimo Village 水下之

郷 in 1565, which explained that both “housemen” (gokenin 御家人) and “document holders” (goinpan-shū 御印判衆) must show up for river maintenance at the Ryūō outpost.103 In this case, housemen referred to military retainers, while document holders meant anyone who had received a confirmation document from the Takeda.

The implication is that all who possessed some kind of investiture from the Takeda owed construction duty as part of their obligation. Later Tokugawa-era documents mentioned above show property tax exemptions in exchange for participation in riparian projects. In these instances, construction duty was treated in the same way military service had been previously – as the basis for retainers’ primary exemptions.

D. Conclusions: Gaining a Stake in the New Order

When we examine Sengoku history as a whole, the new policies of regional warlords loom large as transformative forces. However, the changes that did come to late medieval society were incremental, spread out over several generations, and the result of much trial and error. The basic relationship which developed between the

Takeda and their subjects resulted from various practical solutions to concerns of the moment. Over time, it became apparent to Kai residents that integration within the

102 Shibatsuji et al., Takeda-shi kashindan, 196–97. 103 Yamanashi kenshi shiryō-hen 4, 952.

174

Takeda order, a tacit acknowledgement of their legitimate authority, offered tangible benefits. Though it entailed surrender of autonomy in certain respects, the ability to dispatch a village militia, for instance, the loss of certain kinds of independence proved to be preferable in the long run. It took some time for this to become apparent, and, as we saw with resource disputes, had to be demonstrated.

Although submission ostensibly subjected local residents to a new class of officials, in practice they retained a high level of influence over administration of their areas. The fact that small time landlords were increasingly subjected to Takeda oversight granted ordinary residents considerable leverage. Submission also locked superiors into a set of obligations, foremost of which were backing local claims and maintaining the peace. If those functions were not effectively carried out, the Takeda and their officials essentially lost any claim to authority. So important was this mutual arrangement that by the end of the sixteenth century, rulers gave explicit guarantees of local administrative autonomy. In 1590, the former Takeda retainer Ina Ietsugu 伊

那家次 (unknown) issued a pledge (kishōmon 起請文) to the residents of Kusuji 九筋 village near Kōfu.104 Ietsugu now served Tokugawa Ieyasu, and had been enfeoffed at

Kusuji. Much of his 1590 pledge is very general statements about following rules and doing the right thing, but there are several telling points. In matters including the allocation of tax obligations, exile, tax exemptions due to disaster, and payment of the production tax, Ietsugu stated that all shall be carried out in accordance with the will of the village council (hyakushō-shū no nozomi no gotoku ni 百姓衆のそミのことく

104 Ibid., 295–96.

175

に). In other words, Ietsugu’s role was primarily to sign off on the administrative decisions of the council. Kusuji was left to manage their own affairs for the most part, provided basic obligations were met.

Arrangements such as the one at Kusuji were a continuation of trends initiated in the mid sixteenth century. They grew out of the Takeda desire to promote stability within Kai, and the utility of Takeda influence for local residents in asserting various claims. These relationships continued largely unchanged from the late sixteenth century and into the early seventeenth. Some of the most common types of documents from this period are Tokugawa-issued confirmation letters from 1582 and 1583 asserting that things shall remain as they were before. In many cases, the Tokugawa used the exact same form and wording as previously issued orders. For instance, a

1582 prohibition list (kinsei 禁制) granted to Ichirenji was identical to one issued in

1574 under Takeda Katsuyori.105

Because the late medieval historical record is biased towards documents produced by elites, we must lean heavily upon them to reconstruct a picture of society. Doing so can obscure the fact that real transformation was not enacted by decree or order. These documents reflect settlements that had been worked out over time. Many crucial pieces of that process, the input of people at lower socioeconomic levels, are unfortunately largely lost to history. But each confirmation letter, tax exemption, or prohibition from above was a response to a request. How else could

105 Ibid., 289, 293. Prohibitions were ostensibly regulations, but they also reaffirmed temple rights, for instance, exclusive use of a resource.

176 superiors know what local claims were, or where the border between one field and the next lay? In all the implementation of Takeda and later Tokugawa administration, rulers relied heavily upon local knowledge, and generally endorsed established practices.106 In this way, the very elite and the very local got most of what they wanted from the hierarchical relationship thus established.

The Takeda and other regional warlords predicated their rule on their ability to impose order. We have examined how that could be effective in co-opting local communities to submit to Takeda authority. If the Takeda could not fulfill their role as peacekeeper, however, their power could be undermined rapidly. This situation locked the regional lord into sometimes conflicting obligations. Within Kai, most people desired enough stability to ply their trade and make key decisions locally.

Military retainers, on the other hand, strove for advancement on the battlefield. This group had to be kept satisfied with the rewards of conquest. Autocratic as their power appears, civil war daimyo were thus in a precarious position. To be successful as a civil war daimyo, therefore, was to effectively manage conflicting interests in a way that minimized conflict. The daimyo too, had to act in a way consistent with that necessity. All gained an increased stake in maintenance of stability, and likewise all were subjected to maintenance of it.

Progress towards stability was consistently upset in the civil war era for a variety of reasons. In moments of crisis, tensions were amplified and often erupted in violence. The ensuing chaos could interrupt whatever headway had been made.

106 Philip Brown has argued for precisely this phenomenon in the implementation of the land suvey (kenchi 検地) in the Maeda domain. Brown, Central Authority and Local Autonomy, chap. 3.

177

Gradually, the trend in Kai tipped in favor of maintaining the Takeda-led sociopolitical order. This was a fraught process that took time. But eventually the momentum towards order in Kai outweighed the longstanding civil war tumult. By examining a significant crisis initiated by a natural disaster in 1540, we can see how and when that momentum shifted. I turn now to the 1540 typhoon and Takeda Nobutora’s exile from

Kai.

178

IV. Storming into Exile: Disaster, Embedded Tension, and the

Complex Historical Legacy of Calamity

On a late summer day in the year Tenbun 天文 10 (1541), the warlord of Kai

Province left his son in law’s domain to return home. Traveling with hundreds of men- at-arms, Takeda Nobutora was one of the most powerful people in the Japanese islands.

He controlled a large swath of territory and had many high-placed friends, including his former rival and now son in law Imagawa Yoshimoto 今川義元 (1519-1560), whose domain encompassed neighboring Suruga (see Map x.1). But all of that ended in a single day in 1541. As Nobutora’s entourage made their way north towards home, an even larger cohort of foot soldiers blocked their entry into Kai.1 These troops worked for

Nobutora’s eldest son Shingen. Indeed, Nobutora had been betrayed. The warlord and his loyalists were overwhelmed, powerless to resist a far superior force, and stunned by the treachery of their erstwhile comrades. Hostile warriors marched the Takeda lord right back the way he came, where he was forced seek asylum Yoshimoto in Suruga.

Three days later, on the seventeenth of the sixth month, Nobutora was formally banished from Kai. 2 His twenty one year old son assumed the mantle of power and took over the family mansion at Tsutsujigasaki. Eighteenth generation head of the Takeda house, Nobutora, was finished as a Warring States daimyo. In the span of a few days, a long war-torn territory, populated by thousands of samurai and tens of thousands of battle-hardened residents no strangers to combat themselves, progressed smoothly

1 Kōyō nikki, 86. 2 Ōdai-ki, 353.

179

into the era of a new lord. During an age of near constant strife, somehow the Takeda domain did not descend into internecine war. In fact, under Shingen’s leadership, Kai flourished. How could such a serious leadership crisis be resolved so swiftly amidst such a chaotic age?

It was not unusual for a Warring States daimyo to be betrayed. Retainers, sons, brothers, or uncles might turn coat at any time. It was quite unusual for such a coup to be uncontested. Nobutora’s exile and subsequent resolution must be placed within the context of a larger crisis initiated by a severe natural disaster. This episode was not a straightforward case of retainers replacing an unpopular leader. The lead up to that fateful day in 1541 reveals a complex interplay of tensions embedded within the late medieval socioeconomic hierarchy. Disaster fundamentally shaped how those tensions were expressed, to sometimes surprising results. While it undermined a powerful lord, it simultaneously created avenues for a relatively smooth, non-violent resolution of an ongoing crisis. Disaster here operated as both a destabilizing and stabilizing force. The complex, sometimes contradictory effects of natural disasters make it difficult to evaluate the historical legacy of these events. Discreet causes and impacts are nearly impossible to draw out. However, disaster can be a very effective tool for examining influential tensions and underlying trends present in a given society. In the Takeda case examined here, disaster illuminates a key historical moment when vested interest in maintaining the current sociopolitical order collectively overcame atomization and upheaval. The balance tipped from prevailing disorder towards order in Kai at this time.

Disaster throws the latent tensions ever present in society into sharper relief. It creates conditions that affect the expression of intrinsic tensions profoundly. Dire

180

circumstances produce a sense of urgency which in turn heightens the need to resolve embedded conflicts. These moments often force a settlement of some kind, and as a result rapidly strengthen or weaken various positions on competing sides of an issue.

Tensions that may have long simmered are suddenly thrust into the spotlight. These moments open up possibilities that could not have existed otherwise, while destroying others. The result is often a novel outcome unanticipated by contending parties. During the Tenbun disaster, this phenomenon allowed Takeda Shingen to take over Kai without bloodshed.3

This disaster exacerbated prevailing late medieval tensions, in particular those between retainer and lord, while at the same time it strengthened ordinarily weak bonds between contentious samurai. This unusual juxtaposition resulted from the post- disaster tendency to fortify both prevailing tensions and prevailing bonds within a given society. In other words, natural disaster catalyzed both centripetal and centrifugal tendencies characteristic of the civil war era. In this case, the net effect was actually a reduction of violence.

We cannot impose a linear progression of cause and effect upon analysis of natural disasters. Despite its crucial role in the Tenbun crisis, it would be inaccurate to conclude that disaster caused Nobutora’s exile. Natural disasters tend to catalyze existing trends rather than create lasting structural change.4 With respect to the case under examination, the Tenbun crisis did not fundamentally alter the historical

3 Hereafter I refer to the disaster, its aftermath, and Shingen’s 1541 coup together as either the “Tenbun disaster” or “Tenbun crisis.” 4 Gregory Smits, When the Earth Roars Lessons from the History of Earthquakes in Japan (Rowman & Littlefield Publishers, 2014), xii.

181

developments of the previous few decades in Kai. Rather, this event helped solidify some emerging socioeconomic trends that were already in motion and gaining strength.

The society became more firmly entrenched upon a path towards stability, a path which we have seen relied upon a new, Takeda-led socioeconomic system. That path was already established by 1541. The disaster, therefore, did not bring structural changes. If that is the case here and perhaps true of disaster in general, we might reasonably ask if they are in fact the significant events they appear to be.

Every few years, people around the world are reminded of nature’s tremendous destructive power. In the internet age we hear of almost every greater-than-normal hurricane, tsunami, or tornado anywhere on the globe. These events are obviously significant to those that experience them, that witness them from afar, and those who study them. And yet, there is nothing extraordinary about natural disasters.

Vulnerability to disaster is embedded in our built environments. Indeed, the

“disastrousness” of any natural event is a function of that vulnerability. The term itself

“natural disaster” obscures its human-centric definition. Despite the sense of urgency disaster creates, the long term impact of natural disaster appears negligible. Natural disaster are quite regular historical phenomenon. From that perspective, why do they matter?

Disasters are important historically, but recent scholarship has helped reconceptualize analysis of significance. In his study of earthquakes, historian Gregory

Smits has argued that even these frequent and devastating disasters lack the power to fundamentally alter Japanese society.5 He writes, “insofar as disasters stimulate change,

5 Smits, When the Earth Roars, 162.

182

it is more typically in the manner of a catalyst accelerating processes already underway.”6 Daniel Aldrich, another disaster scholar, has shown how disaster creates conflicting social trends simultaneously. He examines “social capital,” the level of cohesion within a particular group, and argues that it is the most influential factor in determining the post-disaster activity of the group.7 When faced with crisis, a group with high social capital will cooperate more effectively and manage negative circumstances better than a group with low social capital. But this cohesion is simultaneously a tool of solidarity and exclusion. The more resilient a community, the more likely they are to exclude outsiders and even hinder others’ ability to manage a crisis.

Aldrich’s study shows how disaster both increases latent tensions and fortifies underlying social bonds. For example, after the Kansai earthquake in 1923, close knit

Japanese communities at the ward level banded together for mutual support. They also murdered Korean immigrants perceived as threats and denied them access to relief materials.8 After hurricane Katrina struck New Orleans in 2005, communities with high social capital recovered more effectively. They provided mutual support to members, resulting in far lower mortality, less property damage, and faster rebuilding of damaged neighborhoods. They also blocked the siting of temporary housing (the much-maligned

“FEMA trailers”) within their neighborhoods, delaying the recovery of those outside their group.9 At the national level, Katrina sparked debates about racial discrimination.

6 Ibid., xii. 7 Daniel P. Aldrich, Building Resilience: Social Capital in Post-Disaster Recovery (University of Chicago Press, 2012). Esp. chapter 2, “Social Capital: A Janus-Faced Resource for Recovery.” 8 Ibid., 55. 9 Ibid., 145.

183

For contemporary observers of the United States, it is no surprise that race proved one of the most salient tensions exacerbated by this crisis. Even if Katrina recovery was not shaped by racial attitudes as evidence suggests, the important point for our purpose here is that people regarded it as a crucial factor. Race discrimination was the underlying tension came to the fore. Locally, Katrina may have strengthened the trend towards solidarity in already close knit communities. But on a wider scale the hurricane aggravated one of the most longstanding problems in American society.

These and other recent disaster studies have emphasized the limited structural impact and paradoxical social impact of natural disaster generally. But these events can be used as tools to discover a range of influential historical forces. From that perspective, it is not what a disaster changes but how it acts as an agent of change that provides the greatest insight. Disaster reveals deep, structural elements of a society that otherwise might not be visible. These fundamental characteristics manifest in times of crisis, and if examined, are the real key to understanding how events play out. It explains how a rebel warrior could oust an entrenched leader and take over an entire province in the middle of a war without fighting. The outcome of the Tenbun disaster shows that the trend towards reestablished stability had reached critical mass in late medieval Kai.

A. A Hundred Year Storm

184

Typhoon season for central Japan lasts from late summer to mid fall. A number of these storms affect the region to varying degrees every year. But the 1540 season witnessed a storm of exceptional magnitude across huge swath of Honshū. Damage stretched from the chūbu 中部 region (around modern Ōsaka) in the east to the tōhoku

東北 (around modern Fukushima) in the northwest.10 Descriptions of this storm are numerous and dramatic in sources from all over central Japan. Even usually stiff religious chronicles used evocative language to describe the event. In the words of one monk, the extent of destruction was simply “unprecedented in all the world.”11

Kai’s Lake Kawaguchi and proximity (Map 4.1) witnessed the 1540 storm’s fury firsthand. By this time written from the vantage point of Yoshida village lakeside,

Myōhōji-ki vividly depicts the tempest and its aftermath. The following description is based upon that chronicle.12 Monk writers recorded unfavorable weather beginning in the fifth month. There was a deluge and flooding. Some three months later, a massive typhoon hit on the eleventh of the eighth month, September twenty first by the

Gregorian calendar. Winds roared from sundown until the hour of the boar, about ten pm. People living near Lake Kawaguchi were swept away by waves, houses were

10 Shibatsuji Shunroku, ed., Takeda Nobutora no subete (Tokyo: Shin Jinbutsu Ōraisha, 2007), 158. 11 Myōhōji-ki 妙法寺記, 32. Saru hodo ni yo no naka no koto mōshi ni oyobazu sōrō 去程二世中ノヿ申二不及候. 12 Ibid., 31–32. This chronicle uses the word “heavy winds” 大風 to describe major storms, including the one under discussion in 1540. While the word can be read taifū (typhoon), based upon the timing of some storms, it cannot be consistently translated as such. The word should be considered a general term for powerful storms, some of which fall under the modern technical definition of a typhoon. In this case from 1540, the massive storm in the eighth month (September in the Gregorian calendar) was almost certainly a typhoon. The modern word typhoon is written with characters meaning “towering wind” 台風.

185

Lake Kawaguchi Myōhōji

Jōzaiji Lower Yoshida Katsuyama Area Village

Lake Motosu Upper Yoshida Village

Lake Yamanaka

Mt. Fuji

Map 4.1: Lake Kawaguchi and Environs

knocked over, and temples and shrines destroyed. The author wrote of the damage,

“[only] one in one thousand, no, one in ten-thousand commoners’ homes remain

[standing].”13 He goes on to say that there is not a single tree left in the forest, all the animals have been killed, and that some ten-thousand trees in distant Suwa had been knocked over (see Map 4.2).14 The fallout in the following year (1541) appears to have been even worse. Storm damage led to widespread starvation, massive die offs of people and livestock, and a huge number of displaced persons.15 The monk chronicler in

Yoshida wrote that there were “more [displaced] people came [to Yoshida] than there have been in one hundred years;” no small statement to make in the middle of a

13 Myōhōji-ki, 31. Jige no ie wa sen ni, ichiman ni hitotsu goza [sōrō] 地下ノ家ハ千ニ一万ニ御座『候』. 14 While the forest is unspecified, it refers to wooded areas around Yoshida. The assertion that “all the animals have been killed,” and that “ten thousand trees” in Suwa were uprooted are obviously not precise calculations, but rather descriptive expressions of severe damage. 15 Myōhōji-ki, 32–33.

186

decades-long civil war. “One thousand die for each one that lives,” is the author’s grim assessment in 1541.16 People also died of disease. Undoubtedly exacerbated by hunger, exposure, and a suddenly large transient population, epidemic began to claim lives.17

This was a very serious storm.

We can judge the severity of this typhoon by comparing other entries of Myōhōji-ki, and other documents from within and outside of Kai. Myōhōji-ki’s Tenbun 9 entry is considerably longer, more detailed, and more graphic than other descriptions of natural disasters, of which there are many. This was a proverbial hundred-year storm. Other

Kai documents support that conclusion. Ōdai-ki 王代記, also a temple chronicle, and based in central Kai, records strong winds which knocked over trees.18 In Kōfu 甲府,

Takeda retainer Komai Kōhakusai wrote in his journal of heavy winds and damage to the Takeda mansion (see Map 4.2).19 Similar reports appear in the chronicle of the

Rinzai temple Kōgakuji 向嶽寺, slightly northeast of Kōfu in what is today Kōshū City.20

At nearby Daizenji 大善寺, the roof of the main hall collapsed. Inside, images of Nikko

Bosatsu 日光菩薩 and twelve others, all dating back to the early Kamakura era (1185-

1333) were damaged by rain.21 Far from being isolated, this was a massive, multi- regional storm

16 Ibid., 33. Hyakunen no uchi ni mo goza naku sōrō to hitobito mōshikitari sōrō sen shi hitotsu ikiru to mōshi sōrō 百年ノ内ニモ無御座候ト人々申來リ候千死一生ト申候. 17 Kōyō nikki, 86. 18 Ōdai-ki, 353. This chronicle was written from the perspective of the head monk at Hachimansan Fugenji 八幡山普賢寺, an attached temple (bettō 別当) of Kubo Hachiman Shrine 窪八幡宮 in present Yamanashi City. 19 Kōyō nikki, 86. 20 Enzan kōgaku zen’an shōnendai-ki 塩山向嶽禅庵小年代記 in Ibid., 115. 21 Yamanashi kenshi shiryō-hen 4, 616.

187

Lake Suwa

Suwa domain Area of 1528 Takeda/Suwa battle

Suwa Grand Shrine (Upper Main Shrine)

Sakai River Tsutsujigaseki

Daizenji

Kōfu

Takeda domain Map 4.2: Kai-Shinano Border Area

that appeared in numerous across multiple regions of the archipelago.22 Severe damage stretched across the entire middle third of Honshū.23

This major storm wreaked its own havoc, while compounding the pain of other lesser earlier disasters in the region. Punctuated by the massive gale in 1540, the 1530s had been a particularly harsh decade in Kai, even by civil war era standards. Takeda armies maintained an intense campaigning schedule throughout these years.

Concurrently, beginning in Tenbun 3 (1534), every year for the remainder of the decade witnessed some sort of calamity.24 A spate of earthquakes, droughts, epidemics, and famine beset Kai in various combinations for this six year period. The late 1530s were

22 See section 2 of the appendix for a list of these sources and a brief summary of their accounts of the storm. 23 Ikeda Shōichirō, Nihon saihen tsūshi (Tokyo: Shin Jinbutsu Ōraisha, 2004), 311. 24 Some areas, such as Yoshida, had compounding problems virtually for the entirety of the 1530s. Yoshida experienced drought and flooding in the early part of the decade that led to a local water dispute (see chapter 2).

188

probably the most disastrous years for the province in the entire civil war era. But this was not the first time that compounding hardship visited the people of Kai. The waning years of the previous decade had been similarly dire. But reactions to the Tenbun crisis were very different, indicating an important shift in attitudes towards the regional lord had occurred.

B. Medieval Disaster Response and Debt Relief in Kai

Several of Takeda Nobutora’s decisions in the wake of the 1540 storm aggravated tensions between him, his retainers, and many residents of the province. As the sick, starving residents of Kai recovered from the storm, they got no relief from above. The

Takeda lord instead remained focused on his overall geopolitical goals, and continued to campaign in Shinano. Matters only worsened for the people of Kai. Continued mobilization sent troops marching back and forth to the front, and after the typhoon there were “warriors all over the place causing problems.”25 Debt relief, or at least some effort by the ruler to mitigate hardship in the face of this disaster would have been eminently appropriate. His decision to forgo any such measure proved fateful.

Examining why Nobutora chose to forego any kind of disaster relief manifests the inner workings of a fledgling sociopolitical order. Nobutora struggled to balance many competing interests along the socioeconomic spectrum, while simultaneously dealing with inter-regional competition from rival warlords. But, as evidenced by his

(relatively) uncontested removal from power, by the 1540s the new sociopolitical order in Kai was durable enough that the ruler could be replaced without upsetting its basic

25 Myōhōji-ki, 32.

189

functions. This represents an important development in the transition out of the civil war and towards lasting stability.

Ideally, rulers in medieval Japan offered relief measures in times of crisis. Elites demonstrated their wisdom, compassion, and qualifications to govern by enacting policies designed to ease the suffering of the masses. The discourse of the benevolent ruler had a long history in East Asia, and remained prevalent in medieval Japan. By the thirteenth century, the most common official relief policy was the promulgation of special edicts known as tokusei 徳政. Literally “benevolent governance,” the word itself was based upon ancient Chinese precedents. Tokusei evolved in form and function over many centuries, but two aspects remained constant. These measures were designed to ease hardship in times of crisis, and they were one of several obligations a just ruler owed his (the ideal East Asian sovereign was male) subjects.

Originally, “benevolent governance” encompassed a general concept of restoring the proper order of the world after some calamity. Be it a natural disaster, barbarian invasion, or some combination of human and environmental crisis, the state of affairs in the realm reflected upon the ruler’s qualities. When bad things happened it indicated a shortcoming of those in charge, and thus the need for “benevolent governance” in order to restore the proper functioning of society.

Tokusei in Japan evolved from being elite and emperor-centric, to what could be considered populist by the medieval period. The philosophical underpinnings of tokusei date back as far as the eighth century BCE in the Spring and Autumn Annals 春秋, but the first explicit use of tokusei in Japan comes from an 805 CE entry of Nihon Kōki 日本

190

後記.26 Early Japanese tokusei were primarily focused upon the revival or rebuilding of religious institutions. Many commemorated the accession of a new emperor, promulgated as a gift to the populace which displayed the new ruler’s compassion.

Medieval tokusei were quite different. Rather than an abstract notion of restoring righteous order through good government, medieval tokusei consisted of concrete economic policies. In short, tokusei became synonymous with debt relief by the end of the thirteenth century.27 Edicts took on various forms, including partial debt remission, suspension, or wholesale cancellation. Significantly, in the late medieval period, central authorities lost their monopoly on these debt relief measures. “Private” (mizukara 自) and local debt relief (zaichi tokusei 在地徳政) undertaken at the behest of provincial elites, and sometimes communities on their own, began to surface as the central, warrior-led polity declined.28 Incidents of “debt relief rebellions” (tokusei ikki 徳政一

揆), in which commoners forced rulers to cancel debts, have been well-documented by medieval scholars.29 Thus, while ancient tokusei was an imperial prerogative, medieval debt relief was undertaken by provincial magnates, local elites, and the recipients of relief themselves. As part of the effort by civil war warlords to assert exclusive control over their domainal economy, they prevented private and local tokusei.30 Towards the end of the medieval era, more and more economic authority became concentrated in

26 Nihon kokugo daijiten, 9:1149; There is a 751 imperial edict recorded in the Man’yōshū 万葉集 which was functionally a tokusei, though the term was not used. Kokushi daijiten, 10:311. 27 Since tokusei had functionally come to mean debt relief in medieval Japan, hereafter my use of the term “debt relief” indicates a medieval tokusei edict. 28 Seta Katsuya, “Chūsei makki no zaichi tokusei,” Shigaku zasshi 77, no. 9 (September 1968): 1–52. 29 Kokushi daijiten, 10:312. 30 Seta, “Chūsei makki no zaichi tokusei,” 26.

191

the hands of regional magnates. Early modern debt relief was once was again exclusively elite, controlled by daimyo and the Tokugawa bakufu.

A few examples of late medieval debt relief illustrate its evolution from highly elite and abstract to populist and practical. The second warrior-led government in Japan, known as the Muromachi bakufu, first issued a general debt relief edict, meaning it applied to the entire realm, in Kakitsu 嘉吉 1 (1441).31 The measure was a response to unrest and peasant demand for action. That year, intense fighting had taken place in the capital (Kyoto), foreshadowing what was to come in the more protracted Ōnin War several decades later. Kyoto was severely damaged. A war-weakened population also suffered a typhoon and an epidemic.32 On the verge of widespread rebellion, the

Muromachi government opted to cancel debts across the entire archipelago. There are at least four other similar examples of general debt relief by the bakufu from 1441-

1520.33

In the middle decades of the sixteenth century, debt relief became more regionalized. Warlords, not the shogun’s government, had de facto rule of the provinces, and they began to issue debt relief edicts for their “countries.”34 Scholars agree that the

Hōjō 北条 of Sagami Province 相模国 (modern Kanagawa Prefecture) promulgated at

31 Maegawa Yūichirō, “Kabegaki, kōsatsu to Muromachi bakufu tokusei-rei: keishiki kara mita chūsei- hō no kinō,” Shigakku Zasshi 104, no. 1 (January 1995): 2–5. 32 Ikeda, Nihon saihen tsūshi, 280–81. 33 Maegawa, “Kabegaki, kōsatsu,” 5. This only includes countrywide or “general” tokusei, not those issued to specific areas or groups. 34 Indeed, the issuance of tokusei by Warring States daimyo is one of the key indications of their autonomous authority. A general tokusei applied to the entire “country” (in this case a daimyō’s realm of influence) uniformly. Though understood as outward expression of the ruler’s concern and compassion for the subjects of that realm, promulgation of a tokusei implicitly links subjects to the ruler. In doing so, daimyō actively demonstrated that they were the legitimate “public authority” (kōgi) of the domain.

192

least three general debt relief edicts; in Tenbun 19 (1550), Eiroku 永禄 3 (1560), and

Eiroku 4 (1561).35 In each case, the law provided serious concessions to local people in response to a crisis. Each followed war, natural disaster, or both. The Tenbun 19 debt relief cancelled all principal and interest owed, and offered tax breaks for absconded farmers to return to their fields.36 Ten years later in Eiroku 3, commoners demanded debt relief upon the succession of Hōjō Ujimasa 北条氏政 (1538-1590).37 Crops had failed all across the Kantō that year.38 When residents pushed for and won additional debt relief the following year (1561) it was in the midst of a widespread epidemic.39

Here we can see the combination of grass-roots, peasant-driven debt relief with older precedents associating such measures with the succession of a new ruler.

The early 1520s mark a turning point in Kai when Takeda superiority became clear.

At this time, Nobutora issued the a series of investiture documents to a large group of retainers, then toured the province as a display of dominance. These displays of power linked the Takeda to the responsibilities of rulership. Some, such as the overwhelming emphasis upon maintaining the peace, were newly developed from the civil war situation. We have seen how daimyo created a discourse for legitimate rule based upon their ability to mediate and prevent violence within their territories. Other responsibilities of rulership were longstanding traditions that endured throughout the

Warring States era and beyond. Once such continuity was the discourse of benevolent

35 There may have been others, but these three are definite. 36 Noritake Yūichi, “Go-Hōjō ryōgoku-ka no tokusei mondai: eiroku san nen tokusei-rei wo chūshin ni,” Shakai Keizai Shigaku 54, no. 6 (March 1989): 783. 37 The terms of the 1560 edict offered debt cancellation and allowed peasants to pay half of the yearly production tax in rice rather than cash. Ibid., 789; Kuroda Motoki, Sengoku daimyō no kiki kanri (Tokyo: Yoshikawa Kōbunkan, 2005). 38 Fujiki Hisashi, ed., Nihon chūsei kishō saigai-shi nenpyōkō (Tokyo: Kōshi Shoin, 2007), 362. 39 Ibid., 363.

193

governance. Nobutora, warlord conqueror though he may have been, implicitly took on the role of benevolent ruler by asserting Takeda ascendancy in Kai. He became, at least in an abstract sense, responsible for the prosperity of this Takeda-led territory. This was a longstanding feature of East Asian political culture, civil war Japan included.

Nobutora’s own behavior demonstrates that it was influential in late medieval Kai. He issued a general debt relief in 1528 after a string of calamity including natural disaster, economic downturn, and military defeat. This was an even more overt statement that the Takeda ruled Kai, but also that such rulership entailed certain obligations.

The short-term lead up to Nobutora’s first and only debt relief began on the sixteenth of the fifth month of 1528. A heavy rain fell in Kai, resulting in major flooding and “damage to all the fields.”40 Early summer (fifth month in the lunar calendar) is a critical time for growth of rice plants. It is when seedlings have been planted, and stalks are nearing maturation and soon to sprout ears. Around this stage of growth, plants require a gradual reduction of water levels in the fields. In other words, it is not a good time for a flood. But the rains in the fifth month led to just that. On the heels of this major crop failure came a drought lasting from midsummer to early fall (sixth to eighth months). Food prices spiked sharply.41 There was also a plague of insects, probably locusts and probably in early fall, affecting central Kai. 42 It appears 1528 was a bad year for the residents of that province.

Kai warriors had a difficult year as well. Even as the one-two punch of flood then drought wrecked the local economy, Nobutora “mobilized everyone” for a massive push

40 Myōhōji-ki, 21. 41 Ibid., 21–22. 42 Ōdai-ki, 352.

194

into Shinano in early fall.43 His target was the Suwa house, with whom a rivalry had been building for several years. They were warlords themselves, busy carving out a domain in Shinano centered upon their control of Suwa Grand Shrine, a very old and prestigious religious institution. By Eishō 15 (1518) Suwa patriarch Yorimitsu managed to unify roughly one third of the province under his control. In response Nobutora enfeoffed exiled Suwa retainer Kanasashi Masaharu 金刺昌春 (unknown) in Tai’ei 大永

5 (1525), clearly making his intentions towards the Suwa known.44 None of this sat well with Yorimitsu. He was eager to check growing Takeda influence in Shinano. When

Nobutora invaded in the fall of Kyōroku 享禄 1 (1528) Yotimitsu threw his entire might against his foe. The Takeda army lost a pitched battle on the last day of the ninth month along the Kai-Shinano border (Map 4.2).45 It was likely Nobutora’s most costly defeat up to that time.

It was in this atmosphere of distress that Nobutora issued his first and only general debt relief edict in 1528. This measure cancelled debts from the previous three years, and exempted interest payments on new loans for the next three. Other factors could have been involved, but evidence strongly suggests that the edict was in response to the crisis caused by the crop failures in Kai, and the military defeat in Shinano.46

43 Myōhōji-ki, 22. 44 Shibatsuji, Nobutora no subete, 217. 45 Myōhōji-ki, 22. 46 Some scholars speculate that the 1528 era name change from Tai’ei to Kyōroku provided the impetus for Nobutora’s 1528 tokusei, but this raises questions about why this occasion and not others. See Fujiyoshida-shi shi shiryō-hen dai nikan, 238. It is difficult to establish a clear timeline of events since the month and day of the 1528 debt relief are unrecorded. Nonetheless, most scholars conclude that it was indeed a reaction to the dual crisis of famine and defeat. Suzuki Masanori suggests that Nobutora issued the edict as a way to stave off province-wide rebellion in these circumstances. See Yamanashi kenshi tsūshi-hen 2, 273; Suzuki Masanori, “Sengoku daimyō Takeda-shi no tokusei,” Komazawa Shigaku 駒沢史学, no. 74 (2010): 89.

195

Unsurprisingly, food shortages resulted in famine the following year, 1529.47 It is hard to determine whether or not the debt relief edict had any effect, although it is true that the next five years were comparatively prosperous in Kai.48 More importantly, though, is the fact that Nobutora took visible, tangible steps in response to the 1528 crisis. It was a way to mollify a population beset by disaster, large scale mobilization, then defeat.49

Not everyone was thrilled by the debt cancellation, though, and this may in part explain why Nobutora never again issued a tokusei edict. It was unpopular with the creditor and financial class. Myōhōji-ki, itself a product of the financial class, and from an area dependent upon religious tourism to Mt. Fuji, describes there being both

“happiness” and “grief” in response to the debt relief.50 No doubt the money lending monks and wealthy pilgrimage guides considered debt cancellation a cause of their own suffering. In the estimation of the monk chronicler(s), the majority of people “lamented” the measure.51 Considering the social composition of medieval Yoshida, and the perspective of Myōhōji-ki’s writer(s), we can understand the assertion. On the other hand, the passage reveals that a significant portion of Yoshida’s population applauded the tokusei. Even in this relatively moneyed pilgrimage hub, a percentage of residents were debtors. That ratio would have been even higher, and likely reversed, in less commercially viable areas. Province-wide, debt relief probably affected more residents positively.

47 Myōhōji-ki, 23. 48 Ibid., 21–27. 49 Suzuki, “Takeda-shi no tokusei,” 90. 50 Myōhōji-ki, 22. 51 Ibid.

196

Nobutora’s calculus for the 1528 edict was no doubt a difficult one. Hereditary warriors, his most dangerous subordinates, were likely to have varied and volatile reactions to any such sweeping economic proclamations. On the one hand, most people understood tokusei to be an attempt at benevolent government, yet lenders would be as harmed by the measure as borrowers were helped by it. From Nobutora’s perspective, this was likely a risky move either way. Takeda supremacy is Kai was still very young and fragile. Nobutora therefore had to carefully balance the necessities of keeping Kai warriors largely content with keeping them firmly subordinated. Warriors collectively comprised the most politically and economically influential strata of medieval society. It was Takeda retainers, after all, who were ultimately responsible for both Nobutora’s success and his removal from power. But medieval warriors were not a monolithic group, nor was the status clearly defined.52 Their reactions to debt cancellation were as variegated as other socioeconomic groups. Like other groups, most warriors were not elite, and not particularly wealthy. Most possessed a modest amount of wealth and influence in a small area – a single village or less, and at most a handful of settlements.

Like other country folk and city dwellers, they too had debts to pay. Even a relatively affluent warrior might rely on loans to purchase land, build a house, or commission a suit of armor.53 Only the very upper echelon of warrior houses would have possessed the means ability to be major creditors. We can reasonably conclude that a tokusei would have proven quite welcome to the majority of warriors. Nobutora likely based

52 See chapter 1 for a discussion of warriors and the late medieval status system. 53 For a detailed picture of the medieval local economy in English, see Segal, Coins, Trade, and the State, esp. chs. 2, 4, and 5; and Tonomura, Community and Commerce. Segal takes a macroeconomic perspective, while Tonomura’s work considers a particular region.

197

his implementation of the 1528 edict on the necessity to maintain the support of this key demographic.

Even county-level elites in Kai cannot be said to have unilaterally suffered from tokusei. For starters, money lending was generally the domain of religious institutions.

Some influential warriors must have been involved in loans, but not as extensively as temples and shrines. So from a strictly economic standpoint, debt relief was likely a mixed bag for elite medieval samurai. Politically, it was essential. Like the Takeda domain on a smaller scale, county and local elites had their own extensive territories to administer. Most of their subjects, including their samurai retainers, were likely to have benefitted from debt relief. For big-time Kai samurai, Nobutora’s 1528 tokusei was as expedient for them as it was to the Takeda lord. Since the Takeda had established supremacy over Kai’s economic (and other) policy, samurai landlords essentially relied upon the Takeda to provide relief in their own territories.54

Unfortunately, the handful of lines in Myōhōji-ki remains the sole source of information about the effects of the Kyōroku 1 debt relief edict and reactions to it. We simply do not know how most people responded. What can be said for sure is that 1528 was the only time Nobutora provided any kind of relief in response to crisis. The next

Takeda-issued debt relief did not arrive until 1550 under Shingen, a gap of some twenty

54 The level of control exercised over the domain by Takeda warlords remains a matter of some debate amongst scholars. Representing the traditional assessment, Shibatsuji Shunroku considers it to have been quite high, resembling that of an early modern daimyo. Sasamoto Shōji and others have emphasized the high degree of autonomy retained by elite retainers. I am more convinced by the latter position. However, regarding the matter of debt relief, no examples from Kai exist other than those issued by the Takeda. We cannot say for sure that there were no “private” tokusei, or that the Takeda forbade such measures. From the evidence that does exist, it is fair to say that the Takeda took an active economic role. Takeda edicts from the 1520s onwards were applied unilaterally across the domain, even in retainer-held land. For instance, Shingen dictated the amount of tax his retainers could collect from their territories. See Sasamoto, Sengoku daimyō Takeda-shi no kenkyū, esp. chap. 2.

198

two years.55 All told, there were at least six general debt relief edicts in the Takeda domain under Nobutora’s successors, plus two more under Tokugawa rule after 1582.56

The frequency of these edicts and their continued presence indicates the enduring importance of the discourse of benevolent rule.57

Like the late 1520s, the late 1530s were a time of compounding hardship in Kai.

Circumstances grew progressively worse from 1534 onwards, but even in the most desperate of times, Takeda Nobutora took no action to provide relief to struggling retainers and residents. In the midst of a prolonged, draining, military stalemate in

Shinano, the province was beset by famine and epidemic. When the storm hit in 1540, it would have been an opportune time for Nobutora to, outwardly at least, assist his struggling subjects. But he did not, apparently determining that continued focus on the northwestern expansion was more important. There were logical reasons why he reached that conclusion due to the polarizing effect a tokusei was likely to have, yet it proved to be a miscalculation. Nobutora overestimated the support he enjoyed from retainers and his ability to keep his subordinates in line. He did rightly estimate that serious challenges to Takeda supremacy were unlikely. No rival warrior lineage could realistically unseat the Takeda by this point. Doing so would upend recent decades of settlements that, in large part, had helped local people stake a firmer claim to property and helped them resolve local disputes. As examined in chapter 3, such submission came with trade-offs. Nevertheless, the sociopolitical relationships thus forged had

55 Shingen may have issued a general tokusei in 1541 immediately after assuming power. However, scholars disagree over whether or not such an edict existed, a matter I discuss at length below. 56 Suzuki, “Takeda-shi no tokusei,” 86. 57 If anything, it became even more influential in the early modern period as a key ideology of government.

199

coalesced into something preferable than the extreme “self-help” of the civil war by the

1540s.

But Nobutora himself could be replaced. Ultimately he was, largely without incident and without interruption of prevailing, pre-crisis trends. In other words, the system operated independently of ruler himself. This unanticipated outcome reveals significant progress towards stability present in 1540s Kai. It was headed towards a new kind of regularization. By this time, enough people had a vested interest in maintaining the

Takeda-led order that even the serious Tenbun crisis did not amount to a widespread disruption of it. In fact, the crisis actually accelerated developmental processes of that order. We can see how this took place by examining fundamental, latent tensions between socioeconomic groups in late medieval Kai.

C. Embedded Tensions in Late Medieval Society

Disaster response, or rather the lack thereof, ultimately ended the career of a prosperous civil war warlord. Up until his exile from Kai in Tenbun 10, Nobutora had been an extraordinarily successful Sengoku daimyo. Like other warlords, he faced repeated challenges from samurai both within and outside of his domain. But Nobutora had managed to stay on top for over three decades. From 1506 to 1541 he was the head of the Takeda house, and presided over a dramatic expansion of Takeda power and territory. The Takeda managed to survive the upheavals of the late fifteenth century, something that cannot be said of many other warrior houses. Nobutora (unwillingly) bequeathed to his son an expanding, increasingly stable domain. By bringing all of Kai

Province under Takeda sway as an independent territorial lord, he was by far the most

200

successful Takeda leader prior to his son. That was a rare feat during the civil war era.

But his ouster reveals characteristics of three influential dynamics at work in late medieval society. First, it shows how the fundamental conflict of interest between daimyo and provincial warriors was evolving from a contest for supremacy into a new iteration of a center-periphery hierarchy in the archipelago. Second, it reveals the limits of warlord power, even as Warring States magnates, and the constraints the new political system imposed upon superiors. And finally, the Tenbun disaster exhibits a flexible sociopolitical hierarchy which respected local desires and adapted in critical moments. This last characteristic became a defining feature of the early modern order.

In midcentury Kai, it helped usher in the most peaceful coup of the civil war era.

While most people in premodern Japan lacked direct political power, “the ruled” collectively wielded enormous influence. The most obvious examples include direct action such as rebellion or absconding. Hōjō warlords faced both, and responded with debt relief in the 1560s. In other provinces, peasant leagues achieved autonomy and territorial rule rivaling that of warlords.58 Since the ascendancy of the Takeda in Kai in the 1520s, there had been no significant peasant revolts. On the other hand, the countryside was far from being completely pacified. As discussed in chapter 2, it was not until past midcentury that local violence could be kept largely in check. In the late

1530s and early 1540s, mechanisms of local administration were still very much in flux.

In other words, the people of the province remained a serious potential threat to the

Takeda position.

58 The “Single-mind League” (ikkō ikki 一向一揆) is the best example of this phenomenon.

201

Retainers presented a potential threat. Their loyalty, essentially bought off through investiture and war spoils, had a direct correlation to Takeda military and power, and a disproportionate effect on Takeda political influence. Conqueror though he was,

Nobutora had to carefully manage the loyalties of this group.59 It appears he met with mixed results in this regard.60 Throughout his career, Nobutora faced several serious retainer rebellions. Most of these incidents occurred in the context of what was an inherently volatile political moment during the late medieval age: succession.

Even under ideal circumstances, succession is always a potential source of disorder.

During the civil war, smooth succession was noteworthy, particularly amongst warlord houses that controlled large territory and large retainer bands. Given the context,

Shingen’s bloodless, and so far as we know uncontested, takeover of one of the most powerful warlord houses in Japan is quite remarkable. What was unremarkable about the Takeda domain was the conflict of interest between Nobutora and his retainers. The resulting tension between warlords and lower samurai can and often did tear Warring

States domains asunder. The Tenbun disaster certainly aggravated tensions between

Takeda retainers and their lord. These were nothing new. But how those tensions played out in the wake of the storm suggests that the domain had turned a corner towards a more secure political order.

For regional magnates, maintenance of internal cohesion proved just as challenging as defense against external threats. Nobutora himself came to power amdist a

59 See chapter 3. 60 This may have had less to do with his actions per se, and was more a function of chronology. Based on the level of retainer resistance he faced, Shingen appears to have been the more effective ruler. However, Shingen inherited a political system that had been initially forged in Nobutora’s day.

202

succession dispute.61 His ascension to Takeda family headship, though, was quite unlike his son’s. It began when Takeda Nobutsuna 武田信縄 (1471-1507), seventeenth generation Takeda patriarch and military governor of Kai, died in Eishō 永正 4 (1507).

Nobutsuna’s younger brother Aburakawa Nobuyoshi 油川信恵 (d. 1508), leader of a

Takeda branch line who were powerful in their own right, moved to take his elder’s place. It was not the first time Nobuyoshi had attempted to take headship of the main

Takeda lineage. Some fifteen years prior, the two brothers clashed when their father

Takeda Nobumasa 武田信昌 (1447-1505) retired and handed family headship to the elder Nobutsuna in Meiō 1 (1492), but then reneged and granted inheritance to his second son Nobuyoshi.62 Nobumasa’s vacillation eventually resulted in a protracted dispute of various factions and family members. Nobuyoshi and Nobutsuna battled for a year, when the latter finally bested his brother in Meiō 3 (1494). He then overcame his father’s attempt to reclaim family headship after four more years of struggle.63 There was relative stability within the Takeda house for about a decade, but Nobuyoshi made a second bid for family headship in the wake of his elder brother’s 1507 death. This time, his rival was the deceased patriarch’s son Nobutora. The faction supporting fifteen

61 The following summary of events leading to Nobutora’s rise is taken from Yamanashi kenshi tsūshi- hen 2, 262–70. 62 Nobumasa retired and bequeathed headship to Nobutsuna, but then changed his mind only a few months later. He reversed course and endorsed the younger Nobuyoshi, which set off hostilities between the brothers. After Nobutsuna eliminated his brother’s faction in 1494, Nobumasa attempted to remove Nobutsuna by coming out of retirement and reclaiming family headship. Father and son fought until Nobumasa relented and an uneasy truce was reached in 1498. However, after some six years of conflict, factional divisions remained. In 1505, Nobumasa died, leaving Nobutsuna seemingly secure in his position. Only two years later, Nobutsuna died, leaving his eldest son Nobutora to claim headship. But Nobutsuna’s death resulted in a renewed push by the once-defeated Nobuyoshi faction, and Nobutora found himself battling for supremacy against his father’s younger brother. Ōdai-ki, 343; Myōhōji-ki, 6. 63 Myōhōji-ki, 7; Shibatsuji, Nobutora no subete, 213.

203

year old Nobutora emerged victorious against those backing his uncle Nobuyoshi in late

1508. Among those on the losing side were the Oyamada, and Nobutora’s first few years as Takeda head were occupied by conflict against Oyamada, the premier samurai house of southeastern Kai.64

Protracted succession disputes spanning multiple years and generations were unexceptional in civil war era Japan. While Nobutora enjoyed a comparatively long reign as Takeda head, he was frequently involved in the succession disputes of other warrior houses. Perhaps the most significant was Nobutora’s backing of one of four competing Imagawa heirs upon the death of Imagawa Ujiteru 今川氏輝 (1513-1536).65

Rulers of Suruga, to the southwest of Kai, headship of the Imagawa had ramifications for the Takeda. Although Nobutora’s support of the winning Imagawa faction initiated a new era of alliance between the rulers of Kai and Suruga, his involvement came with a price. The Suruga succession dispute led to the severe punishment of a Takeda retainer house and a serious backlash against Nobutora.

Takeda-Imagawa conflict had a long history. Suruga warriors first invaded Kai in

Eishō 11 (1514), and continued to do so on a more or less regular basis despite attempts at peace. One ambitious attempt concluded in Eishō 14 (1517) when Nobutora married the daughter of Ōi Nobusato 大井信達 (unknown), an Imagawa ally. Nobusato’s daughter Ōi no Kata 大井の方 (1497-1552) became Nobutora’s principle wife and later

64 Shibatsuji, Nobutora no subete, 214. 65 Although Ujiteru’s assumption of Imagawa headship appears to have been smooth, it was only possible because of his savvy and influential mother, Jukeini 寿桂尼 (d. 1568). One of the most remarkable women of the civil war era, she effectively ruled the province for two years from her husband Imagawa Ujichika’s 今川氏親 (1471-1526) death until Ujiteru’s sixteenth birthday in 1528. See Kubota Masaki, “Imagawa Ujichika kōshitsu Jukeini hakkyū no monjo ni tsuite,” Komazawa Shigaku 24 (1976): 135–48.

204

the mother of Shingen. The marriage came on the heels of a major Imagawa defeat at

Katsunuma, in central Kai, by the Kobayashi of Yoshida. The Kobayashi of course used their fighting prowess to increase their standing in Yoshida, serving as local military commanders (yorioya) for several decades.66

Despite the marriage of Nobutora and Lady Ōi, lasting peace between Takeda and

Imagawa remained elusive. As allegiances seesawed back and forth in the tumult of war, on one occasion Nobutora found himself simultaneously targeted by his two most dangerous foes. The Hōjō and Imagawa put aside their own differences in Tenbun 4

(1535) and raided Kai. Yoshida village, victim of a Hōjō assault only four years prior, bore the brunt of the invasion. Both the upper and lower sections of the settlement were put to the torch. 67 Imagawa-Hōjō cooperation was a nightmare scenario for

Nobutora. Fortunately for him, architect of that alliance and Imagawa patriarch, Ujiteru, died the year after Yoshida was burned. Ujiteru’s death ended the joint invasion of Kai, and thrust Suruga into internal chaos. The late patriarch’s four younger brothers, all of whom had been living as monks, vied for family headship in a conflict known as the

Hanakura Rebellion (Hanakura no Ran 花蔵の乱). One was a young acolyte at Fujigun

Zentokuji 富士郡善得寺 named Zengakujōbō 栴岳承芳, who left monastic life and took the name Yoshimoto.68 Yoshimoto ultimately bested his brothers, largely thanks to Hōjō intervention. While Nobutora did not actively aid Yoshimoto, he did decide to endorse

66 See chapter 2. The Kobayashi defeat of Imagawa occurred during a period of growing Kobayashi influence in Yoshida. They had sponsored major infrastructure projects in the early sixteenth century (1504), and their successful defense of Yoshida in 1517 very likely contributed to the family’s increased material and political clout. This ultimately led to increased tensions between them and other Yoshida residents by midcentury. 67 Myōhōji-ki, 29. 68 Shibatsuji, Nobutora no subete, 72.

205

the new leader of Suruga. When he offered his eldest daughter Jōkei’in 定恵院 (1519-

1550) in marriage in 1536, the alliance was sealed.

Takeda Nobutora finally had his long-sought truce with Suruga, after many years and no small amount of blood spilled. But his retainers were not as committed to this newfound friendship with the ascendant Imagawa lord. One house, the Maeshima 前島, sheltered a group of Yoshimoto’s enemies named Kushima 福島 in Kai. The Kushima were veterans of the Kai-Suruga war, and had opposed Yoshimoto during the Hanakura

Rebellion. They fled after his victory. Eager to secure the goodwill of his new ally

Yoshimoto, Nobutora ordered the Kushima extradited to Suruga. Nobutora then when a step further, indicating the high priority he placed upon the Kai-Suruga truce. He ruthlessly ordered the Maeshima to commit suicide en masse as punishment for harboring the enemy.69 This action went against wishes of a significant number of influential Kai samurai. Takeda retainers reacted swiftly and decisively. A large group of retainer-administrators renounced Nobutora and fled the province in protest.70 Some ended up in Sagami as Hōjō retainers.71 Outside of Shingen’s coup, it was the most serious rejection Nobutora faced during his tenure.

Succession disputes notwithstanding, the conditions of late medieval Japan provided ample opportunities for conflict between lord and retainer. Nobutora, like other warlords of the day, rose to prominence by subduing competing warrior houses

69 Myōhōji-ki, 29; Kōyō nikki, 84. 70 Myōhōji-ki, 29. The retainers are referred to as “magistrates” (bugyō 奉行), which is the standard translation employed for Edo period Japan. While the exact function of late medieval magistrates in the Takeda domain is unclear, it appears to refer to retainers holding some sort of administrative office. 71 Shibatsuji, Nobutora no subete, 73.

206

in his home province. He was a conqueror. Takeda rule of Kai in the civil war era is often characterized as a transition from military governor to Warring States daimyo, but this obscures the means by which Takeda dominance continued post Zenshū’s

Rebellion. Kai did not belong to the Takeda or any other warrior house for most of the fifteenth century.72 Any one of a half dozen or so houses contended for supremacy.

Nobutora owned a prestigious name, and that certainly did not hinder his ability to gather supporters. But Kai became the Takeda realm under Nobutora because his armies wrested it away from competitors. Naturally, some vanquished foes-turned- retainers harbored resentment. Even under ideal circumstances warlords struggled to keep their samurai in check. Nobutora’s rise and reign were littered with rebellions and challenges from his own followers. By comparison, Shingen may have enjoyed more good will simply because he inherited the realm rather than conquered it.

Whatever the case, after defeating his uncle and taking over from his father,

Nobutora’s leadership was typical of a civil war warlord. His spent the first few years in power fighting to maintain his position vis-à-vis county and local warriors. Nobutora’s subjugation of the Oyamada in 1510 was his first major victory. 73 It solidified his position as Takeda patriarch. For the remainder of the Eishō era (1504-1520)

Nobutora’s armies regularly faced internal threats even as they began to venture beyond provincial borders in efforts to expand Takeda territory. Battles with kokujin in

72 For example, when Takeda Nobshige, Nobutora’s great-great grandfather, was appointed military governor of Kai by the Muromachi shōgun in 1421, the province was in such disarray that he refused to leave Kyoto to take up the post. See chapter 1. 73 Myōhōji-ki, 12.

207

southwestern Kai and against the Imagawa became a regular feature of the first decade under Nobutora.74

The Takeda house gradually gained supremacy over county and local warriors. In a tangible expression of burgeoning Takeda dominance, Nobutora moved the Takeda seat to the centrally located Kōfu in 1516. A coalition of Kai warrior houses pushed back. 75

But within two years the Takeda had broken their opposition. The 1521 gathering at

Shōhōji, where Nobutora issued confirmation letters to warriors under his fealty is considered by most scholars to mark the unification of Kai under Takeda rule.76 Others, including Hirayama Masaru 平山淳, cite continued rebellions against Takeda authority in the 1520s and argue that Kai should not be considered the “Takeda realm” until as late as 1532.77 However, Nobutora’s implementation of the property tax in 1522 indicates that the Takeda had indeed achieved ascendancy at this time.78 The tax was not popular, but nonetheless by this time Nobutora was able to require it of his vassals.

While the 1520s did not mark the end of internal challenges for the Takeda, it does represent a turning point. Post 1521 military campaigns tended to be externally focused, for territorial expansion. It is true that county and local warriors occasionally rose up against the Takeda during this period. But the relationship between lesser samurai houses and the Takeda had qualitatively changed. Conflicts within Kai were no

74 Most major kokujin families in Kai were branch lineages of the Takeda, some from many generations past. This group included the Hemi 逸見, Imai 今井, Ōi 大井, Anayama 穴山, and Kurihara 栗原 warrior houses. When Nobutora came to power, the Ōi and Anayama were Imagawa followers. Thus, some one third of Kai Province was under Imagawa sway at the time. Yamanashi kenshi tsūshi-hen 2, 265–66. 75 In this conflict, anti Takeda forces consisted of the Hemi, Ōi, Kurihara, and Itagaki houses. 76 Yamanashi kenshi tsūshi-hen 2, 271. 77 Hirayama Masaru, Takeda Shingen (Tokyo: Yoshikawa Kōbunkan, 2006), 9; Shibatsuji, Nobutora no subete, 219. 78 See chapter 3 for a discussion of the Shōhōji meeting and the property tax (munabetsu) in Kai.

208

longer multi-way scrums in the civil war power vacuum. Rather, they became rebellions by inferiors against their superiors.

Despite Takeda preeminence, Nobutora’s position was by no means guaranteed.

Warlords had to keep their retainers happy. During the civil war, the way to do that was to campaign and distribute rewards to followers. Nobutora’s major external campaigns during the second and third decades of his rule (1520s-1530s) were against the Hōjō to the southeast, and the Suwa to the northwest. By that time, the Hōjō had become the major warrior house of the Kantō plain, replacing the old bakufu officials from the

Uesugi lineage. In Shinano, the Suwa held sway over roughly one third of the province, in the southeast corner bordering Kai. Both regions offered more lucrative agricultural resources than were to be had in Kai. After repeatedly being stymied by the Hōjō, expansion towards the northeast appeared more attractive to Nobutora. Important as it was to continually dole out rewards to retainers, Nobutora’s strategy seems to have been to carve away at Suwa territory bit by bit. He planned to install the Kanasashi, a branch line of the Suwa as well as former retainers, as heads of Suwa Grand Shrine.

They had joined the Takeda vassal corps after their exile from Shinano in 1525, and were happy to be Takeda deputies there if it meant a restoration of the family’s fortunes.79 By 1528, the stage was set for a major Takeda-Suwa showdown.

Nobutora’s Shinano invasion would prove to be pivotal for his tenure as ruler. At the first battle against the Suwa in Kyōroku 1 (1528), the Takeda suffered a significant defeat along the Kai-Shinano border near the Sakai River 境川 (see Map 4.2).80 As

79 Yamanashi kenshi tsūshi-hen 2, 272. 80 Myōhōji-ki, 22.

209

discussed above, that same year Nobutora issued his first and only general debt relief.

The defeat foreshadowed what would be a largely unsuccessful push to the northwest for the Takeda. Subsequent forays into Shinano met with mixed results, which was probably the worst possible scenario for the Takeda leader. While a decisive victory or defeat would have given Nobutora a clear course of action, lukewarm success prolonged the conflict with little tangible benefits. Nobutora’s Shinano invasion could be labeled a quagmire in contemporary terms. His opponent was not a single, powerful warlord like himself or his Imagawa and Hōjō rivals. Rather, in the 1530s, Shinano remained disunified, controlled by county and local warriors with little or no overall coordination.81 Sometimes these warrior houses came together to resist Takeda advances. But their coalitions arose and dissolved as circumstances changed. Since there was no single foe to beat, there was no singular victory to be achieved. Conquest and subjugation were piecemeal, labor-intensive processes. This appears to have hindered Nobutora’s plans for expansion into Shinano. Despite the challenges, Nobutora remained committed to expanding his northwestern flank, despite evidently uneven support from his vassal band.

Nobutora avoided significant vassal revolt after the Suwa defeat and disaster in

1528. But an active anti-Nobutora faction formed in Kyōroku 4 (1531). At the time, the

Hōjō were once again aggressive on Kai’s southeastern border after several quiet years.

Also threatened by Hōjō expansion were the Ōgigayatsu Uesugi 扇谷上杉, one of two

Uesugi lines that had shared the post of Eastern Deputy under the Muromachi

81 Shibatsuji, Nobutora no subete, 14.

210

shoguns.82 The Uesugi steadily lost territory to Hōjō incursions from the mid 1520s, and their leader Tomo’oki 朝興 (1488-1537) had proposed an alliance with the Takeda as early as 1525. It finally materialized in 1531. At Tomo’oki’s behest, Nobutora took deceased Yamanouchi Uesugi 山ノ内上杉 patriarch Norifusa’s 上杉憲房 (1467-1525) widow as a concubine. 83 Some Takeda retainers opposed the move. To make matters worse, the Ōgigayatsu-Takeda alliance never paid dividends. Hōjō troops burst into the religious and trading village of Yoshida on the twenty third of the fourth month in 1531, killing many.84 But no military cooperation between Takeda and Ōgigayatsu ever materialized.85 After the defeat, tensions mounted within the Takeda retainer band.

Nobutora responded by forcing the heads of the Obu 飯富 and Kurihara 栗原 families, both high ranking, into retirement.86 Another retainer, Imai Nobumoto 今井信元 (1484-

1575) then gathered a group of like-minded, anti-Nobutora samurai and allied with

Suwa Yorimitsu 頼満 (1473-1540), Nobutora’s old enemy in Shinano. This set off a widespread series of loosely connected rebellions in both Kai and Shinano.87 After four

82 See chapter 1. 83 Unfortunately, little is known of this woman. She appears in documents only as “Uesugi Tomomasa’s 上杉朝昌(unknown) daughter.” Norifusa held the post of Eastern Deputy, an office Tomo’oki desired for himself and his own Ōgigayatsu line of the Uesugi. After Norifusa died, Tomo’oki gained influence over both Uesugi lineages, and tension between them largely died down. 84 Myōhōji-ki, 23. 85 Despite this, Nobutora deepened his ties with Tomo’oki. Takeda heir Shingen married Tomo’oki’s daughter, a woman whose name remains unknown, in 1533. She died in childbirth the following year. Local tradition has it that she and Shingen never got along. Like a large part of well-known Takeda lore, there is no historical evidence to support this story. What is important here is that Nobutora doubled down on a policy (alliance with Uesugi) that many of his retainers disliked. 86 Myōhōji-ki, 24. The family heads were Obu Toramasa 飯富虎正 (1504?-1565) and Kurihara Hyōgo 栗原兵庫 (d. 1531). Obu Toramasa is believed to have been instrumental in Takeda Shingen’s coup, and went on to become one of his most trusted retainers. 87 Ibid.

211

years of on-again off-again fighting, Nobutora concluded a shaky truce with Yorimitsu and company in Tenbun 4 (1535).

Relative peace along Kai’s borders offered little respite. As discussed above, the late

1530s were some of the most trying times for the region. Amidst the myriad calamities of those years, the Takeda ended up back at war with the Suwa four years after the uneasy peace in 1535. Suwa Yorimitsu died in Tenbun 8 (1539), leaving his grandson

Yorishige 頼重 (1516-1542) to rule in his own right. 88 The new leader had ambitions to retake territory lost to the Takeda. He reversed his predecessor’s nonaggressive stance and marched against the Takeda, allied with the Ogasawara, another influential house in central Shinano. Armies clashed all through 1540, beginning in the fifth month and continuing even after the typhoon hit in early fall.89 Disaster victims were forced to deal with hordes of samurai marching to and from the front in Shinano.

Nobutora’s Shinano campaign had been ongoing for some twelve years all told.

Takeda forces campaigned in the midst of the Tenbun disaster, as starvation and disease ravaged their homeland. Despite this monumental effort, Nobutora’s followers had little to show for it. Three months after the momentous storm, the Shinano invasion ended in decidedly anti-climactic fashion. In the eleventh month of 1540, Nobutora negotiated the marriage of his third daughter Nene 禰々 (1528-1543) to Suwa

Yorishige. A truce thus concluded, the Takeda-Suwa border remained as it was in 1528.

Lack of records means we can only speculate, but it is not hard to imagine that many

88 Yorimitsu’s heir Yoritaka 頼隆 (1499-1530) unexpectedly died young on the eighteenth of the fourth month in 1530. Three years later Yorimitsu officially retired and took the tonsure so that he could pass family headship to his grandson Yorishige and ensure a smooth succession. Though Yorishige officially held the position of family head, grandfather Yorimitsu ruled in fact until his death in 1539. 89 Myōhōji-ki, 32.

212

Takeda retainers felt Nobutora had mishandled the Shinano invasion. He had chosen to campaign through the 1540 disaster, a move that was clearly unpopular with local people in Kai.90 And negotiated peace left little material rewards to be doled out to ambitious warriors. It likely did little to ease tensions between lord and retainer.91

Like many other civil war warlords, Takeda Nobutora developed a tense relationship with his heir as well. While clear evidence is sparse, scholars have speculated that Nobutora and Shingen clashed due to philosophical differences in ruling style. Shingen appears to have styled himself a Confucian ruler. He was at least familiar with the Chinese classics, and presented himself as a wise and just leader befitting that tradition. During Shingen’s tenure, the Takeda law code for retainers, Kōshū hattō no shidai 甲州法度之次代, drew upon classical Chinese law. It was likely written by

Shingen’s younger brother Takeda Nobushige 武田信繁 (1525-1561), himself a student of the classics.92 Whether from philosophical differences or not, Nobutora reportedly came to prefer the younger Nobushige as heir. In a famous anecdote from Kōyō gunkan

甲陽軍艦, a hagiography of Shingen, Nobutora refused to toast his eldest son and heir at a New Year’s party in Tenbun 7 (1538).93 The slight was very public and very clear in its message. Evocative as the story is, it may not be true. What we can say for certain is that

90 Myōhōji-ki, 32. We do not have records of individual retainers’ reactions. However, as most were themselves local samurai, it is reasonable to conclude that their reactions to the 1540 disaster were largely in line with the rest of the population. 91 Nobutora’s leadership during the Shinano invasion can be contrasted with that of his successor. Discussed in detail below, Shingen suspended campaigning for a whole year as Kai dealt with the impact of the 1540 storm. He then restarted the Shinano invasion and brought it to a successful conclusion, eliminating the Suwa. This essentially added Shinano to Takeda territory. This appears to have been widely supported by Takeda retainers. 92 Shibatsuji, Nobutora no subete, 15. 93 Hirose Kōichi, Akaoka Shigeki, and Kai sōsho kankōkai, eds., Kai sōsho 甲斐叢書, vol. 4: Kōyō gunkan, jō, (Tokyo: Daiichi Shobō, 1974), 32.

213

there was at least some level of tension between Nobutora and Shingen. Whether or not

Nobutora intended to disinherit his eldest in favor of Nobushige cannot be asserted with confidence. Regardless, the bond between brothers turned out to be stronger than that between father and sons. Nobushige remained loyal to Shingen throughout his life, becoming a trusted adviser and one of his brother’s top generals.94 He died on the field at the famous fourth battle of Kawanakajima 川中島 in Eiroku 4 (1561).

Nobutora was demonized in Kai after his exile, as longstanding grievances against the former warlord became legitimate, or even appropriate, overnight. Monks in

Yoshida attributed his untimely departure as retribution for “excessive evil deeds.”95

The precise nature of these evil deeds was never specified. While embedded tensions against Nobutora were given free expression, Shingen, the new Takeda lord, was praised within Kai and condemned from outside. Arch-rival Uesugi Kenshin 上杉謙信

(1530-1578), wrote of the coup as an evil, unfilial act. Never mind that only weeks before Nobutora too had been Kenshin’s enemy. But such an about face was the nature of civil war society. The rhetoric of friendship, loyalty, service, and alliance was constantly adjusted to fit the practical needs of the moment.

In Kai, praise for Shingen and denouncement of Nobutora resolved the outstanding problems of the Tenbun crisis. Reaction to the coup in existing historical sources was overwhelmingly positive. Takeda retainers celebrated the ascension of their new

94 We can only speculate, but if Nobutora had indeed intended to disinherit his eldest son, Nobushige’s later devotion to his brother takes on a slightly different significance. As his brother’s potential usurper, Nobushige may have realized the importance of demonstrating his loyalty after Shingen assumed power. 95 Myōhōji-ki, 32.

214

leader.96 Commoners and samurai alike lauded the beginning of the “age of lord

Shingen,” and there was “unprecedented happiness” for those high and low. 97 Although unsurprising that the new Takeda leader would be praised at the expense of his predecessor, the fact that there was no serious opposition to Shingen’s coup is noteworthy.98 The coup and subsequent harmonious endorsement of Shingen met the needs of the moment. Even at the local level, there was recognition that Nobutora had mishandled the Tenbun crisis, and his replacement was appropriate. Analysis of the disaster context of the coup reveals that there was much more to the story than rebellion against an unpopular ruler, led by an ambitious heir. It also explains how, contrary to generations of observed samurai behavior, Takeda retainers united amidst a contested succession. Then Tenbun disaster triggered a sudden crisis, while simultaneously providing avenues for a swift, almost clean, resolution.

D. Implications – Do Disasters Matter?

Shingen’s coup was a remarkable event for late medieval Japan. On the one hand, a natural disaster undermined an extremely successful regional ruler. The devastation precipitated a serious political crisis, to say nothing of the widespread, on-the-ground hardship. This catastrophe ultimately resulted in regime change. And yet, quicker than it arose, the crisis was resolved. An entire province wrested away from its warlord with barely a sword unsheathed. How could this once in a century disaster inspire both

96 Kōyō nikki, 86. 97 Myōhōji-ki, 32, 33. 98 It is certainly possible that some people privately opposed Shingen, however, no such records exist. Outwardly, Shingen enjoyed the full support of the retainer band and residents of Kai, no small feat during the Sengoku period.

215

rebellion and an unprecedented concord amongst local warriors? How could such a seemingly cataclysmic event affect so much and yet leave so much unchanged? Did the disaster actually have an impact on these events? The region’s political and administrative structure remained largely intact. No initiatives for better preparedness, or more effective disaster relief, emerged after 1541. Indeed, life went on much as it had before for people at all levels of society.

While the Tenbun disaster did not result in structural change, it helped solidify a number of important trends. Initial chaos aside, in the long term it increased momentum towards stability in late medieval Kai. The crisis stressed a developing, relatively fragile political order and provided an opportunity for unhappy subordinates to replace their leader. But it also created an avenue for a tidy outcome to an inherently volatile situation. The fact that retainers coalesced around Shingen, elevating him to the same position and role his predecessor occupied, can be considered a kind of litmus test for the durability of the sociopolitical order midcentury. Enough people had a stake in maintaining this order that it survived a serious shock. As argued above, that vested interest developed only after people saw how integration could help them solve their local problems. In 1540, that process was still underway. There were, as yet, no guarantees the province would not disintegrate into chaos. But, as we saw in the 1533 water dispute between Yoshida and Watanabe, the Takeda-led ability to resolve local conflicts and maintain stability was slowly increasing in local areas. Yoshida’s feud with

Watanabe, only seven years prior to the Tenbun crisis, shows that local people had begun to embrace a new “public authority” (kōgi).

216

The fact that nothing substantively changed for the people of Kai after Shingen took over shows that administrative functions were becoming more regularized. The rhetoric that Shingen was radically different from his father comes almost exclusively from Kōyō gunkan and local tradition. In historical sources considered to be more reliable, it is hard to detect significant differences in the ruling styles of Nobutora and his son. Even if we consider that Shingen was more heavily influenced by Chinese classics, it is hard to connect that to any appreciable effects on his policies.99 Militarily,

Shingen continued his father’s campaign of expansion into Shinano to the northwest.

Only one year after Nobutora’s exile, Takeda warriors were right back at war with the

Suwa, fighting against the armies of Yorishige.100 Politically, there was no major shakeup of the retainer band, and of course no drastic change in the way Kai was ruled.

Shingen, as his father before him, controlled territory by securing the loyalty of a class of professional warriors. Local administration remained unchanged. Over time, governance became more streamlined, more centralized, and more effective in Kai, but that was no radical departure from what had been established during Nobutora’s time.

Shingen’s tenure was a continuation and expansion of a process initiated under his father and mirrored in other warlord domains across the archipelago.

The only significant change between Nobutora and Shingen occurred in the immediate aftermath of the disaster. Shingen may have actually issued a debt relief edict upon his succession in the summer of Tenbun 10 (1541). A 1541 entry from the chronicle Ōdai-ki reads “ikkoku heikin anzen 一国平均安全” in reference to changes

99 Shingen did of course adopt the slogan “Fūrinkazan 風林火山” on his war banners, an acronym derived Sun Tzu’s classic. 100 Myōhōji-ki, 33.

217

wrought by Shingen’s ascension. Some historians, including Hirayama Masaru and

Kuroda Motoki believe this to be reference to a general debt relief edict.101 If that assessment is true, it paints an evocative picture of the contrast between Shingen and

Nobutora. The new Takeda leader offered a domain-wide debt relief measure in the face of unprecedented hardship – the very thing Nobutora failed to do. Other scholars are not convinced that the phrase actually refers to debt relief. Suzuki Masanori argues that this sentence is merely an acknowledgement that Shingen suspended military campaigning during the 1540-1541 crisis.102 In either case, this indicates recognition of tangible responses on the part of the ruler to the suffering of his subjects. That was the important thing. Exhibiting the form of a just, worthy ruler was far more important than anything effective Nobutora could have done in the wake of the typhoon.

Nobutora could have been removed from power under very different circumstances.

It would seem that all of the key ingredients had been in place for some time: an ambitious heir who had run afoul of his father, simmering resentment amongst a significant percentage of the retainer band, mediocre success on the battlefield, as in the late 1520s and 30s, and several spates of natural disasters exacerbating it all. Takeda retainers could have replaced their leader on any number of occasions had there been a critical mass willing to do so. 103 But any attempt to get rid of Nobutora before 1541 would have very likely resulted in factional war. We cannot know if those in Shingen’s

101 Hirayama, Shingen; Kuroda Motoki 黒田基樹, Sengoku-ki no saimu to tokusei (Tokyo: Azekura Shobō, 2009). The phrase ikkoku heikin 一国平均 is often used in various documents as a way to describe tokusei edicts. However, the entry in Ōdai-ki does not reference debts or finances explicitly, but merely says anzen 安全 (“peace” or “safety”). Given that other Takeda tokusei edicts are more explicitly described, this does not appear to be a debt relief edict. 102 Suzuki, “Takeda-shi no tokusei,” 91. 103 Shibatsuji, Nobutora no subete, 161.

218

coup consciously made that choice. Most likely, retainers acted in their own pragmatic interests. At that critical moment, the quick ouster of an established leader and consent around a new one made the most sense. It was a very practical, and self-centered, decision, yet one that actually helped preserve order. It is very telling that retainers’ self-interest coincided with maintaining the Takeda-led sociopolitical system.

Shingen’s treatment of his father after the coup suggests that his motivation was pragmatic rather than personal. For his part, he did not hold a grudge. He made arrangements with Yoshimoto for Nobutora to be given a stipend, a modest estate, and for his wives and servants to be sent to join him in Suruga.104 Theories that this correspondence indicated a secret arrangement with Yoshimoto, as is suggested in Kōyō gunkan, have been discredited by medieval scholars.105 From Yoshimoto’s perspective, the arrangement was a definite plus. Although he had an existing alliance with Takeda

Nobutora in 1541, the opportunity to forge a new peace with the much younger Shingen represented a longer term solution to Kai-Suruga relations. It also gave him potential leverage, at least initially. Had Shingen been unsuccessful, or had Kai deteriorated into internecine struggle after the coup, Yoshimoto could have sponsored Nobutora’s return to power and ruled Kai as Nobutora’s patron. Moreover, Nobutora provided insurance for Yoshimoto’s alliance with the younger Takeda. If Shingen had truly disliked his father and wished to destroy him, it is unlikely he would have made arrangements for

104 These arrangements are references in a letter from Yoshimoto to Shingen dated to the 23rd day of the 9th month of 1541. Yamanashi-ken shihensan iinkai, ed., Yamanashi kenshi shiryō-hen 5: chūsei 2 ge, kengai monjo (Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 2005), 140. Interestingly, Nobutora’s principle wife and Shingen’s mother Ōi no Kata 大井の方 (1497-1552) remained in Kai, possibly at her son’s behest. She remained a person of significant influence. Her villa in Kōfu was the site of several conferences between high ranking Takeda retainers, and dignitaries from outside warlord domains as recorded in Kōyō nikki. 105 Shibatsuji, Nobutora no subete, 162, 165.

219

Nobutora’s retirement in Suruga. This illustrates that Shingen’s coup was a reaction to a very specific set of circumstances. From this perspective, the 1540 typhoon presented

Shingen a golden opportunity; exacerbating tensions between his father and Takeda retainers, while simultaneously unifying them into a faction of his own.

During the civil war era, sociopolitical fragmentation meant that most groups had very low social capital, to borrow Aldrich’s phrase. People were forced, out of self- preservation, to act prioritize individual needs. This meant that cooperation tended to be very small-scale; at the level of a single household, a cluster of residences, or a village. In one sense, the process of rebuilding a functional order was essentially expanding the scope of cooperation. Conflict continued to exist, but it became regulated by a set of relationships that allowed for complimentary interests between groups to be realized. Probably the most difficult group to integrate in such a way were warriors. In the mid sixteenth century, the Takeda retainer band was an unstable conglomeration with very low social capital. That is to say, they were generally unlikely to act cohesively or make accommodations for a common cause. Warriors were generally selfish, and ruthless. And yet, it simply was not in most retainer’s interest to upend the established order in the 1540s. The opportunity was there in the Tenbun crisis, but stability prevailed.

In terms of impact, what disasters they reveal is far more fundamental than what they change. In this case, it is no surprise that tension between military retainers and their lord came to the fore in a Warring States domain. But the resolution to the Tenbun crisis points to a much more significant shift in late medieval society. Momentum towards stability finally outweighed tendencies towards factionalism, political

220

dysfunction, and violence. Disaster can be a tool to discover this kind of significant historical evolution.

221

Conclusion: Perspectives on Environment and Order

This study has examined the role of environment in large scale social, economic, and political transformation. The sharp contrast between endemic instability and lasting order in late sixteenth and early seventeenth century Japan provides an opportunity to analyze structural causes that contribute to order and disorder in society. One of my primary goals has been to demonstrate that conflicts over local environmental management were key sites of discord in late medieval Japan. These contests represented crucial stakes for those involved, and on the whole mattered far more to most islanders than wars for territory. They were the main stakes of war. As such, inconsistency and endemic violence in environmental disputes had a large cumulative impact. For overall stability to once again prevail in the Japanese islands, there had to be effective, reliable, non-disruptive means of resolving local environmental problems.

We have traced how such mechanisms developed in late medieval Kai Province. The result combined a high degree of local autonomy and local management of resources with an authoritarian political hierarchy that could effectively mediate and keep the peace. This somewhat unlikely combination worked because, as discussed in chapter 3, new relationships across the sociopolitical spectrum resulted in the formation of complimentary interests. These relationships were hierarchical, but not dominated by one side. They included a kind of submission to certain terms for all involved. The new elite, regional magnates, assumed the responsibilities of guarantor, mediator, peace keeper, local booster, and fiscal manager. Local communities acknowledged the lord’s

222 prerogatives and surrendered a portion of their productive output because the daimyo’s government served the public good. But these terms were not abstract for contemporaries. Local people came to accept daimyo authority as it was demonstrated to be a useful tool. The terms of submission were largely secondary to the fact that subordinates gained an influential patron who could back their local claims and help them end local conflicts favorably. From the perspective of the Takeda and other regional lords, the terms of local administration were largely secondary to the fact that they were able to tap into local productivity, and that their rule was recognized as the legitimate “public authority” of the realm. This system evolved similarly in other provinces, and eventually evolved into a political structure in which the outward expression or performance of submission was far more important than functional control.1

The two pillars civil war era territorial states were a productive economy and a strong military. Military retainers comprised the army, but could also be a serious threat. Daimyo consistently worked to keep their retainers happy, usually by confirming landholding rights and granting new ones as rewards for meritorious battlefield service. However, the vast majority of domainal residents were not heredity vassals. It was the productivity of these people upon whom rulers depended for their accumulation of surplus. Daimyo were keen to maintain stability and thus leave local production unimpeded. Warlords developed a rather laisses-faire approach to local economy and society. So long as local communities met their obligations (taxes, labor),

1 For a summary of this phenomenon in early modern Japan, see Roberts, Performing the Great Peace.

223 they were largely left alone. This arrangement fulfilled the basic needs of both territorial rulers and local people. As we have seen from Takeda examples above, when the interests of local communities collided with that of a military retainer, the lord usually restrained the vassal and endorsed local claims. This was part of the process of subjugating the warrior class which Birt and others have described. For the daimyo, the most expedient means of administering territory was to largely preserve local autonomy, while gradually curtailing the autonomy of hard to control warriors.

The stakes of war for most of the people in medieval Kai involved local resources. If we only focus upon the grand geopolitical struggle taking place at the regional level and beyond, the significant, transformative impact local dispute resolution will be missed.

The examples presented from late medieval Yoshida indicate a significant evolution took place over the course of the sixteenth century. In these instances, Takeda lords did not expand their authority by directly claiming large chunks of local resources. Rather, it was their ability to effectively manage these disputes which solidified their position as the new ruling elite. By providing tangible advantages to local residents, the Takeda gained a tacit acknowledgement of their legitimate authority. They accomplished this by demonstrating an ability to protect local claims and respect local precedent. This method of local administration continued to be influential throughout the Edo period.

In the final chapter, I examined a novel outcome to serious natural disaster which struck Kai Province while the major evolution into a new kind of sociopolitical order was underway. This disaster caused a crisis that resulted in Takeda Nobutora’s removal from power, but it was far from a simple case of an unhappy retainer band rebelling

224 against an unpopular lord. The Tenbun crisis functioned as a catalyst for trends towards stability – trends that were established via the dispute resolution mechanisms discussed in chapter 2 and the emerging complimentary interests of different social classes examined in chapter 3. In this moment of crisis, the prevailing tendencies of late medieval Kai society came to the fore, which resulted in a quick, undisruptive, and largely uneventful coup. The fact that such a coup could take place, in particular in the context of long era of civil war, indicates that the society had turned a corner towards the reestablishment of stability. Or to put it another way, this shows a stronger, more durable vested interest in the developing sociopolitical order progressing in the mid sixteenth century.

Through the processes examined in this study, a new balance was struck which eliminated endemic violence and unleashed a wave of economic and demographic growth. Japan’s late medieval to early modern transition has yet to be fully examined from the perspective of the local environment. This study has hopefully demonstrated that there are various productive avenues of research which will contribute to a better understanding of this dynamic age. Yet several important questions remain at the conclusion of this study. Perhaps most obvious, we might ask how typical was Kai

Province? This project has not intended to set Kai up as a bellwether for late medieval

Japan. Rather, it is precisely in the peculiarities of time and place that we can perceive broader historical significance. My Kai-centric Warring States timeline will hopefully contribute to a clearer sense of the key characteristics of this era. Nonetheless, it will be useful to examine local resource disputes, development of complimentary interest, and indications of trends towards stability in other regions. The closely related and

225 influential neighboring provinces around Kai are a logical starting point. We might take this analysis further forward chronologically as well. Although I have repeatedly referred to the influential changes of the early modern age, these developments have not been analyzed here. How did the achievement of stability affect local environmental management? I have only hinted at generalities based upon the current research. We know that the early seventeenth century witnessed rapid and pronounced urban growth. The economy expanded, transforming from the regional, mercantilist spheres which independent domains attempted to manage into a highly integrated, interregional commercial network. Commodification, cross-regional interdependence, and the sheer volume of trade grew to levels never before experienced.

The birth of Japan’s early modern commercial economy has long attracted scholarly attention. However, with a few notable exceptions, studies have centered upon urban consumption, long distance trade networks, the business practices of the merchant class, and elite economic policy. Surely these developments had an impact upon the local environment. This is a question historians only recently began to consider. Karen

Wigen, for example, has connected these influential forces into a rich regional history and elucidate their impact upon a particular landscape and the people who moved through it.2 But is yet much to be explored. Economic studies of the period have already demonstrated a marked increase in specialization. This in turn affected the management of local resources, as preferences for certain types of materials, lands, and products shifted. It appears that local resources became more intensively used, and

2 Karen Wigen, The Making of a Japanese Periphery, 1750-1920 (Berkeley: University of California Press, 1995)

226 rights to those resources increasingly divided.3 Not unlike the layered rights of the old

Estate system, more and more separate groups were forced to share the same productive spaces and economic and demographic demands stressed traditional management practices. Eventually, localized management increasingly gave way to outside commercial forces. A new class of industrialists were able to lease the rights to local resources while local people themselves increasingly relied upon imports for their daily necessities.4 Control of local resources began to be based upon investment, rather than proximity and occupation. This likely drastically altered the way that many communities interacted with the local environment. Examination of the birth and growth of Japan’s commercial economy from the perspective of the local environment can be a useful tool for contemporary society as we grapple with issues of stratification, globalization, commodification, and sustainability.

3 Conrad D. Totman, The Green Archipelago: Forestry in Preindustrial Japan (Berkeley: University of California Press, 1989), 157–58. 4 Totman, The Green Archipelago Conrad D. Totman, The Lumber Industry in Early Modern Japan (Honolulu: University of Hawaii Press, 1995). The financial elite included wealthy merchants and domain financial managers who sometimes oversaw large commercial enterprises with the goal of improving the domain’s economic resources. For the increased cooperation between the merchant class and political elite in the later Tokugawa period, see Luke Shepherd Roberts, Mercantilism in a Japanese Domain: The Merchant Origins of Economic Nationalism in 18th-Century Tosa (Cambridge: Cambridge University Press, 1998).

227

Bibliography

Adolphson, Mikael S. The Gates of Power: Monks, Courtiers, and Warriors in Premodern Japan. Honolulu: University of Hawai’i Press, 2000.

Adolphson, Mikael S., and J. Mark Ramseyer. “The Competitive Enforcement of Property Rights in Medieval Japan: The Role of Temples and Monasteries.” Journal of Economic Behavior and Organization 71 (2009): 660–68.

Aldrich, Daniel P. Building Resilience: Social Capital in Post-Disaster Recovery. University of Chicago Press, 2012.

Amino Yoshihiko 網野善彦. Muen, kugai, raku: Nihon chūsei no jiyū to heiwa 無縁、公界、楽：日本中世の自由と平和. Tokyo: Heibonsha, 1987.

———. Rethinking Japanese History. Translated by Alan S. Christy. Ann Arbor: University of Michigan Press, 2012.

Batten, Bruce L., and Philip C. Brown, eds. Environment and Society in the Japanese Islands: From Prehistory to the Present. Corvallis: Oregon State University Press, 2015.

Berry, Mary Elizabeth. The Culture of Civil War in Kyoto. Berkeley: University of California Press, 1994.

Birt, Michael P. “Warring States: A Study of the Go-Hōjō Daimyo and Domain, 1491- 1590.” PhD diss., Princeton University, 1983.

———. “Samurai in Passage: The Transformation of the Sixteenth-Century Kanto.” Journal of Japanese Studies 11, no. 2 (July 1985): 369–99.

Brown, Philip C. Central Authority and Local Autonomy in the Formation of Early Modern Japan: The Case of Kaga Domain. Stanford: Stanford University Press, 1993.

———. Cultivating Commons: Joint Ownership of Arable Land in Early Modern Japan. Honolulu: University of Hawaiʻi Press, 2011.

Conlan, Thomas. State of War: The Violent Order of Fourteenth-Century Japan. Ann Arbor: University of Michigan Press, 2003.

Eason, David Anthony. “The Culture of Disputes in Early Modern Japan, 1550-1700.” PhD diss., UCLA, 2009.

Farris, William Wayne. Heavenly Warriors: The Evolution of Japan’s Military, 500-1300. Cambridge: Harvard University Press, 1992.

228

———. Japan’s Medieval Population: Famine, Fertility, and Warfare in a Transformative Age. Honolulu: University of Hawai’i Press, 2006.

———. Japan to 1600 : A Social and Economic History. Honolulu: University of Hawaii Press, 2009.

Friday, Karl F. Samurai, Warfare and the State in Early Medieval Japan. Warfare and History. New York: Routledge, 2004.

Fujiki Hisashi 藤木久志. Toyotomi heiwarei to sengoku shakai 戦国の作法: 村の紛争解決. Tokyo: Tokyo Daigaku Shuppankai, 1985.

———. Sengoku no sahō: mura no funsō kaiketsu 豊臣平和令と戦国社会. Tokyo: Heibonsha, 1987.

———, ed. Nihon chūsei kishō saigai-shi nenpyōkō 日本中世気象災害史年表稿. Tokyo: Kōshi Shoin, 2007.

Goodwin, Janet R. Selling Songs and Smiles: The Sex Trade in Heian and Kamakura Japan. Honolulu: University of Hawai’i Press, 2007.

Fujiyoshida-shi kyōiku iinkai 富士吉田市教育委員会, ed. Fujiyoshida-shi shi shiryō sōsho 4, mura meisaichō 富士吉田市史資料叢書 4 村明細帳. Fujiyoshida-shi: Fujiyoshida-shi kyōiku iinkai, 1988.

Fujiyoshida-shi shihensan iinkai 富士吉田市史編さん委員会, ed. Fujiyoshida-shi shi shiryō-hen dai nikan (kodai, chūsei) 富士吉田市史、史料編第二巻古代・中世. Fujiyoshida-shi: Fujiyoshida-shi, 1992.

———, ed. Fujiyoshida-shi shi shiryō-hen dai sankan (kinsei ichi) 富士吉田市史、史料編第三巻近世 I. Fujiyoshida-shi: Fujiyoshida-shi, 1994.

Fujiyoshida-shi shihensan-shitsu 富士吉田市史編さん室, ed. Myōhōji-ki 妙法寺記. Fujiyoshida-shi: Fujiyoshida-shi kyōiku iinkai, 1991.

Hall, John W., Nagahara Keiji, and Yamamura Kōzō, eds. Japan Before Tokugawa: Political Consolidation and Economic Growth, 1500-1650. New Jersey: Princeton University Press, 1981.

Hall, John Whitney. Government and Local Power in Japan, 500 to 1700. Princeton: Princeton University Press, 1966.

229

———. “Terms and Concepts in Japanese Medieval History: An Inquiry into the Problems of Translation.” Journal of Japanese Studies 9, no. 1 (Winter 1983): 1– 32.

Hattori Harunori 服部治則, and Shimizu Shigeo 清水茂夫, eds. Takeda shiryōshū 武田史料集. Tokyo: Jinbutsu Ōraisha, 1967.

Hirayama Masaru 平山優. Takeda Shingen 武田信玄. Tokyo: Yoshikawa Kōbunkan, 2006.

Hirose Kōichi 廣瀨廣一, Akaoka Shigeki 赤岡重樹, and Kai sōsho kankōkai 甲斐叢書刋行會, eds. Kai sōsho yon: Kōyō gunkan, jō 甲斐叢書 4: 甲陽軍艦、上. Tokyo: Daiichi Shobō, 1974.

Ikeda Shōichirō 池田正一郎. Nihon saihen tsūshi 日本災変通志. Tokyo: Shin Jinbutsu Ōraisha, 2004.

Ikegami, Eiko. The Taming of the Samurai: Honorific Individualism and the Making of Modern Japan. Cambridge: Harvard University Press, 1995.

Kokuritsu rekishi minzoku hakubutsukan 国立歴史民俗博物館, ed. Nihon shōen shiryō 日本荘園資料. Tokyo: Yoshikawa Kōbunkan, 1998.

Kokushi daijiten henshū iinkai 国史大辞典編集委員会, ed. Kokushi daijiten 國史大辭典. Yoshikawa Kōbunkan 吉川弘文館, 1979.

Kubota Masaki 久保田昌希. “Imagawa Ujichika kōshitsu jukeini hakkyū no monjo ni tsuite 今川氏親後室寿桂尼発給の文書について.” Komazawa Shigaku 駒澤史学 24 (1976): 135–48.

Kuroda Motoki 黒田基樹. Sengoku daimyō no kiki kanri 戦国大名の危機管理. Tokyo: Yoshikawa Kōbunkan, 2005.

———. Sengoku-ki no saimu to tokusei 戦国期の債務と徳政. Tokyo: Azekura Shobō, 2009.

Maegawa Yūichirō 前川祐一郎. “Kabegaki, kōsatsu to Muromachi bakufu tokusei-rei: keishiki kara mita chūsei-hō no kinō 壁書・高札と室町幕府徳政令 : 形式からみた中世法の機能.” Shigakku zasshi 史學雜誌 104, no. 1 (January 1995): 1–33.

Marushima Kazuhiro 丸島和洋. Sengoku daimyo no gaikō 戦国大名の「外交」. Tokyo: Kōdansha, 2013.

230

Mass, Jeffrey P. The Kamakura Bakufu: A Study in Documents. Stanford: Stanford University Press, 1976.

———. The Development of Kamakura Rule, 1180-1250: A History with Documents. Stanford: Stanford University Press, 1979.

———, ed. The Origins of Japan’s Medieval World: Courtiers, Clerics, Warriors, and Peasants in the Fourteenth Century. Stanford: Stanford University Press, 1997.

Matsudaira Sadamasa 松平定能, ed. Kai kokushi 甲斐国志. Tokyo: Yūzankaku, 1968.

McNeill, William H. Plagues and Peoples. New York: Anchor Press, 1976.

Morris, Ivan. The Nobility of Failure: Tragic Heroes in the History of Japan. Tokyo: Kurodaha Press, 1975.

Murdoch, James, Isoh Yamagata, and Joseph H. Longford. A History of Japan. Yokohama: Kelly & Walsh, 1903.

Nagahara Keiji, and Kōzō Yamamura. “Shaping the Process of Unification: Technological Progress in Sixteenth- and Seventeenth-Century Japan.” Journal of Japanese Studies 14, no. 1 (January 1988): 77–109.

Nihon Daijiten Kankōkai 日本大辞典刊行会, ed. Nihon kokugo daijiten 日本國語大辞典. Tokyo: Shōgakukan, 1972.

Nishikawa Kōhei 西川広平. “Sengoku-ki Kai no kuni ni okeru zaimoku no chōsetsu to yamazukuri 戦国期甲斐国における材木の調達と山造.” Takeda-shi kenkyū 武田氏研究 39 (2009): 1–20.

———. Chūsei kōki no kaihatsu: kankyō to chiiki shakai 中世後期の開発・環境と地域社会. Tokyo: Kōshi Shoin, 2012.

Noritake Yūichi 則竹雄一. “Go-Hōjō ryōgoku-ka no tokusei mondai: Eiroku san nen tokusei-rei wo chūshin ni 後北条領国下の徳政問題 : 永禄三年徳政令を中心に.” Shakai keizai shigaku 社會經濟史學 54, no. 6 (March 1989): 780–99, 893.

Osano Asako 小佐野浅子. “Takeda ryōgoku-ka ni okeru kokujin Oyamada-shi to gunnai no zaichi seiryoku 武田領国下における国人小山田氏と郡内の在地勢力.” Takeda-Shi Kenkyū 武田氏研究 30 (2004).

Roberts, Luke Shepherd. Mercantilism in a Japanese Domain: The Merchant Origins of Economic Nationalism in 18th-Century Tosa. Cambridge: Cambridge University Press, 1998.

231

———. Performing the Great Peace: Political Space and Open Secrets in Tokugawa Japan. Honolulu: University of Hawai’i Press, 2012. Sakai Kenji 酒井憲二, ed. Kōyō gunkan taisei 甲陽軍鑑大成. 7 vols. Tokyo: Kyūko Shoin, 1994.

Sano Shizuyo 佐野静代. Chū-kinsei no sonraku to suihen no kankyōshi: keikan, seigyō, shigen kanri 中近世の村落と水辺の環境史: 景観・生業・資源管理. Tokyo: Yoshikawa Kōbunkan, 2008.

Sasamoto Shōji 笹本正治. Sengoku daimyō Takeda-shi no kenkyū 戦国大名武田氏の研究. Kyoto: Shibunkaku Shuppan, 1993.

———. Zenkōji no fushigi to densetsu: shinkō no rekishi to sono miryoku 善光寺の不思議と伝說: 信仰の歴史とその魅力. Nagano-shi: Isshōsha Shuppan, 2007.

Segal, Ethan Isaac. Coins, Trade, and the State: Economic Growth in Early Medieval Japan. Cambridge: Harvard University Press, 2011.

Sengoku Daimyo Jiten Henshū Iinkai 戦国人名辞典編集委員会, ed. Sengoku jinmei jiten 戦国人名辞典. Tokyo: Yoshikawa Kobunkan, 2006.

Seta Katsuya 瀬田勝哉. “Chūsei makki no zaichi tokusei 中世末期の在地徳政.” Shigaku zasshi 史学雑誌 77, no. 9 (September 1968): 1–52.

Shibatsuji Shunroku 柴辻俊六. “Sengoku-ki no suiri kangai to kaihatsu 戦国期の水利灌漑と開発.” Minshū-shi kenkyū 民衆史研究 11 (1973): 79–98.

———. Sengoku daimyō-ryō no kenkyū: Kai Takeda-shi ryō no tenkai 戦国大名領の研究: 甲斐武田氏領の展開. Tokyo: Meicho Shuppan, 1981.

———, ed. Takeda Nobutora no subete 武田信虎のすべて. Tokyo: Shin Jinbutsu Ōraisha, 2007.

Shibatsuji Shunroku 柴辻俊六, Hirayama Masaru 平山優, Kuroda Motoki 黒田基樹, and Marushima Kazuhiro 丸島和洋, eds. Takeda-shi kashindan jinmei jiten 武田氏家臣団人名辞典. Tokyo: Tokyodo Shuppan, 2015.

Shibatsuji Shunroku 柴辻俊六, Kuroda Motoki 黒田基樹, and Marushima Kazuhiro 丸島和洋, eds. Sengoku ibun: Takeda-shi hen 戦國遺文：武田氏編. 6 vols. Tokyo: Tōkyōdō shuppan, 2002.

Smits, Gregory. When the Earth Roars: Lessons from the History of Earthquakes in Japan. Rowman & Littlefield Publishers, 2014.

232

Spafford, David. A Sense of Place: The Political Landscape in Late Medieval Japan. Cambridge: Harvard University Press, 2013.

Suzuki Masanori 鈴木将典. “Sengoku daimyō Takeda-shi no tokusei-rei 戦国大名武田氏の徳政令.” Komazawa shigaku 駒沢史学, no. 74 (2010): 84–97.

Takagi Shōsaku. “Hideyoshi’s Peace and the Transformation of the Bushi Class: The Dissolution of the Autonomy of the Medieval Bushi.” Acta Asiatica: Bulletin of the Institute of Eastern Culture 49 (1985): 46–77.

Takeuchi Rizō 竹內理三, ed. Shōen bunpuzu 荘園分布図. Tokyo: Yoshikawa Kōbunkan, 1975.

Tashiro Hiroshi 田代博志. “Chūsei kōki no shōen sonraku ni okeru funsō kaiketsu to ryōshu 中世後期の荘園村落における紛争解決と領主.” Kumamoto Shigaku 熊本史学 83–84 (2004): 79–108.

Tonomura, Hitomi. Community and Commerce in Late Medieval Japan: The Corporate Villages of Tokuchin-Ho. Stanford: Stanford University Press, 1992.

Totman, Conrad D. The Green Archipelago: Forestry in Preindustrial Japan. Berkeley: University of California Press, 1989.

———. The Lumber Industry in Early Modern Japan. Honolulu: University of Hawaii Press, 1995.

Varley, H. Paul. The Ōnin War; History of Its Origins and Background with a Selective Translation of the Chronicle of Ōnin. New York: Columbia University Press, 1967.

Wigen, Karen. The Making of a Japanese Periphery, 1750-1920. Berkeley: University of California Press, 1995.

Yamamoto Takeshi 山本大, and Owada Tetsuo 小和田哲夫, eds. Sengoku daimyō kashindan jiten: tōgoku-hen 戦国大名家臣団事典：東国編. Tokyo: Shin Jinbutsu Ōraisha, 1981.

Yamamura Kōzō. “From Coins to Rice: Hypotheses on the Kandaka and Kokudaka Systems.” Journal of Japanese Studies 14, no. 2 (July 1988): 341–67.

The Cambridge History of Japan. Vol. 3. Cambridge: Cambridge University Press, 1990.

Yamanashi Kenshi Hensan Iinkai 山梨県史編さん委員会, ed. Yamanashi kenshi shiryō- hen 4: Chūsei 1, kennai monjo 山梨県史資料編 4：中世 1、県内文書. Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 1999.

233

———, ed. Yamanashi kenshi shiryō-hen 6: chūsei 3 jō, kennai kiroku 山梨県史資料編 6：中世 3 上、県内記録. Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 2001.

———, ed. Yamanashi kenshi shiryō-hen 5: Chūsei 2 ge, kengai monjo 山梨県史資料編 5：中世 2 下、県外文書. Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 2005.

———, ed. Yamanashi kenshi tsūshi-hen 2: Chūsei 山梨県史通史編 2：中世. Kōfu-shi: Yamanashi Nichinichi Shinbunsha, 2006.

234

Appendix Section 1: Graphics of Medieval Political Economy

235

236

Section 2: Sources from Outside Kai Describing the Typhoon of 15401

Eastern Japan Name Location Origin/Author Summary Tōdera hachimangū Aizu 会津, southwest Chronicle of Strong winds in the nagachō 塔寺八幡宮 Mutsu Kokoroshimizu evening; old and 長帳 Hachiman Shrine large trees toppled; wind ripped up the earth Shōhō nenpu jūzan- Southeast Mutsu Chronicle of Shōhōji Heavy rain and wind ki 正法年譜住山記 Raikōji nendaiki 来 Western Dewa Chronicle of Raikōji (12th day of the 8th 迎寺年代記 month) powerful wind Imamiya saishi-roku Shimotsuke (14th day of the 8th 今宮祭祀録 month) heavy winds; pagoda at Tōshōji 東勝寺 toppled; damage to temples all over; people blown away; many old trees within shrine grounds felled Akagiyama nendaiki Kōzuke Heavy winds; main 赤城山年代記 hall at Iwadono 岩殿 damaged; famine across several provinces Suwa shinshi ontō Suwa, central Chronicle of Suwa Wind uprooted grass; no nikki 諏訪神使御 Shinano shrine head priests Pagoda at upper 頭之日記 (Moriya 守矢) shrine fell over; large trees beyond count toppled Nendaiki hagō-shō Musashi (various) anthology Many references to 年代記配合抄 heavy winds Kaigen sōzu-ki 快元 Kamakura Journal of the monk Heavy winds, large 僧都記 Kaigen 快元 (b. 1487) trees toppled over or at Tsurugaoka snapped; damage to Hachimangū shrine grounds; severe damage to a number of temples; foliage destroyed;

1 This chart was compiled from the following: Fujiki, Nihon chūsei kishō saigai-shi nenpyōkō, 348– 349; Ikeda, Nihon saihen tsūshi, 311; and Shibatsuji, Nobutora no subete, 157–158. All sources listed explicitly reference the date as the 11th day of the 8th month of Tenbun 9 unless otherwise noted. Entries recorded as one or two days later likely refer to the same storm.

237

unprecedented flooding Shugo nendaiki 守護 Echizen Major flooding 年代記

Western Japan Rokuon nichiroku 鹿 Kyoto Chronicle of Rokuon- Heavy 苑日録 in, sub-temple of winds/typhoon; (12th Shōkokuji, initiated day of the 8th month) by Keijo Shōrin 景徐 treasures of 周麟 (1440-1518) Gokusenji 玉泉寺 damaged by wind and rain; temple roofs and pillars toppled Kyōroku nengo kinai Kyoto Heavy winds/ heiran-ki 長享年後機 typhoon, Go-Nara 内兵乱記 emperor aggrieved by the calamity; typhoon; epidemic across the realm, many deaths, emperor is worried; (10th, 11th, and 13th of the 8th month) Nijōji shuke-ki 二条 Kyoto Heavy rain and wind; 寺主家記 pagoda at Yamashirō Hachiman shrine knocked over; damage to many other places; Kasuga shrine buildings collapsed Dokushi gushō Kyoto Anthology of Heavy chūhen 読史愚抄中 documents compiled winds/typhoon 編 #47 in 新訂増補国史大系 Oyu-dono no ue no Kyoto Compilation of Heavy wind and rain nikki 御湯殿上日記 women court attendants’ journals at old imperial palace Genjo ōnen-ki 厳助往 Kyoto Journal of the monk Heavy wind/typhoon 年記 Genjo at Daigoji 醍醐 across the entire 寺 realm Mokudai nikki 目代 Kyoto Chronicle of Kitano (13th day of the 8th 日記 Tenmangū month) wind and rain; many trees snapped

238

Ten’neiji Tanba Chronicle of (12th day of the nendaiki 天寧寺年代 Ten’neiji 8th month) flooding 記 Kumano nendaiki Kii Chronicle of (8th month, no 熊野年代記 Kumano Hayatama day specified) heavy Grand Shrine winds, flooding; all the rivers flooded and people were swept away and drowned

239